Characterizing rechargeable batteries on the server side

Abstract

The present invention discloses a method for characterizing a rechargeable battery (91-96) on a server side, the method comprising: obtaining operating values ​​for the capacity of the battery (91-96) and the impedance of the battery (91-96); based on the operating values: performing at least one state prediction (181-183) of the battery (91-96), wherein each state prediction of the at least one state prediction (181-183) comprises a plurality of iterations (1099), wherein in each iteration (1099) a simulation of the electrical state of the battery (91-96) and the thermal state of the battery (91-96) is performed and an aging assessment for the capacity and the impedance is determined based on the results, wherein the aging assessment from a first iteration (1099) of the corresponding state prediction (181-183) is used for the simulation in a subsequent second iteration (1099) of the corresponding state prediction (181-183).

Description

Technical Field

[0001] Various embodiments of the present invention are directed to techniques for characterizing batteries. Various embodiments of the present invention are directed to techniques for characterizing at the server side to obtain aging assessment within a prediction interval. Background Art

[0002] Rechargeable batteries, such as power batteries for electric vehicles, have a limited service life. For example, the so-called state of health (SOH) of the battery may decrease over time, for example as a result of charging cycles. This may be problematic, for example in relation to the reliable operation of the corresponding battery-driven device. The SOH is usually determined in combination with the capacity and impedance of the battery.

[0003] Some techniques are known to determine the SOH of a battery in the actual state. For example, such a determination can be made by a battery management system (BMS) of the battery. However, the SOH of a battery in the actual state is only partially convincing as to how reliably the battery can be used in future operation. Summary of the invention

[0004] There is therefore a need for improved techniques for characterizing rechargeable batteries. In particular, there is a need for techniques that can provide reliable information about the future operation of the battery.

[0005] This object is achieved by the features of the independent claim. Embodiments are defined by the features of the dependent claims.

[0006] The method for characterizing a rechargeable battery on the server side comprises obtaining evaluation values ​​for the capacity of the battery and for the impedance of the battery. In addition, the method also comprises performing at least one state prediction of the battery based on these evaluation values. Here, each state prediction of at least one state is associated with a corresponding operating boundary condition of the battery. In addition, each state prediction of at least one state prediction comprises a plurality of iterations. Here, a simulation of the electrical state of the battery and the thermal state of the battery is performed in each iteration and an aging evaluation for the capacity and the impedance is determined based on this result. The aging evaluation resulting from the first iteration of the corresponding state prediction is then used for the simulation in the subsequent second iteration of the corresponding state prediction.

[0007] The computer program or computer program product or computer-readable storage medium includes a program code. This program code can be loaded and run by a processor. This prompts the processor to execute a method for characterizing a rechargeable battery on the server side, wherein the method includes obtaining evaluation values ​​for the capacity of the battery and for the impedance of the battery. In addition, the method also includes performing at least one state prediction of the battery based on these evaluation values. Here, each state prediction in at least one state prediction is associated with the corresponding operating boundary conditions of the battery. In addition, each state prediction in at least one state prediction includes multiple iterations. Here, a simulation of the electrical state of the battery and the thermal state of the battery is performed in each iteration and an aging assessment for the capacity and impedance is determined based on this result. The aging assessment obtained by the first iteration of the corresponding state prediction is then used for the simulation in the subsequent second iteration of the corresponding state prediction.

[0008] The server comprises a processor configured to load a program code and implement the above-described method for characterizing a rechargeable battery on the server side based on the program code.

[0009] The features mentioned above and those yet to be explained below can be used not only in the respectively explicitly stated combination but also in other combinations or alone, without departing from the scope of protection of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 schematically illustrates a system including a plurality of batteries and a server according to various examples;

[0011] Figure 2 showing details of a battery according to various examples;

[0012] Figure 3 showing details of a server according to various examples;

[0013] Figure 4 A flow chart of an exemplary method is schematically shown;

[0014] Figure 5 schematically illustrates results of an aging assessment for multiple state predictions for characterizing a battery according to different examples;

[0015] Figure 6 is a flow chart of an exemplary method, wherein, Figure 6 Some details of simulation of the electrical and thermal states of a battery according to different examples are shown;

[0016] Figure 7 schematically illustrates a collective operating configuration of a battery according to various examples;

[0017] Figure 8 schematically illustrates an event-related operating configuration of a battery according to various examples;

[0018] Fig. 9 schematically illustrates a method for determining operating values ​​of impedance and capacity of a battery according to various examples;

[0019] Fig.10 schematically shows a current time series with a jumping window for determining an operating value of an impedance of a battery according to various examples;

[0020] Fig.11 Standby voltage-state of charge characteristic curves and corresponding compression according to different examples are schematically shown. DETAILED DESCRIPTION

[0021] The above-mentioned characteristics, features and advantages of the present invention and the manner in which these characteristics, features and advantages are achieved will be more clearly understood in conjunction with the following description of the embodiments explained in more detail with reference to the accompanying drawings.

[0022] Next, the present invention will be explained in more detail based on reference to the accompanying drawings and preferred embodiments. In the accompanying drawings, the same reference numerals mark the same or similar elements. The accompanying drawings are schematic representations of different embodiments of the present invention. The elements shown in the accompanying drawings do not need to be shown in correct proportion. More specifically, the different elements shown in the accompanying drawings are reproduced so that those skilled in the art can understand their functions and general uses. The connections and couplings between the functional units and elements shown in the accompanying drawings can also be implemented as indirect connections or couplings. The connections or couplings can be implemented wired or wirelessly. The functional units can be implemented as hardware, software, or a combination of hardware and software.

[0023] Several techniques are described below in connection with the features of a rechargeable battery. The techniques described herein can be used in connection with a wide range of different battery types, such as lithium-ion based batteries, such as lithium-nickel-manganese-cobalt oxide batteries or lithium-manganese oxide batteries.

[0024] The batteries described herein can be used in various fields of application, for example in devices such as motor vehicles or drones or portable electronic devices such as mobile communication devices. It is also conceivable to use the batteries described herein in the form of static energy storage devices.

[0025] The technology described here enables the characterization of a battery based on the prediction of the battery state. This means that the battery state can be predicted within a certain prediction interval in the future. In particular, the battery SOH can be evaluated for aging within the prediction interval.

[0026] As a general rule, the SOH decreases as the battery ages. Aging may be exacerbated if the capacity of the battery decreases and / or the impedance of the battery increases.

[0027] This state prediction can be used in different ways. For example, the result of the state prediction can be used to adjust the operation of the battery. This can, for example, avoid excessive aging of the battery. Alternatively or additionally, the secondary use of the battery can also be determined based on the predictive state. For example, it can be determined whether the battery is still suitable for a specific use or must be withdrawn from circulation. Another possible application involves targeted maintenance and / or repair of the battery based on the state prediction. For example, batteries whose SOH is predicted to deteriorate can be maintained or replaced.

[0028] As a general rule, the state prediction techniques described in this article can be used at different levels: In a simple implementation, the state prediction can be performed at the battery level, that is, a comprehensive aging assessment can be performed horizontally across all battery blocks and battery cells. But in other examples, the state prediction can be performed in an analytical way for different battery blocks or even for the cells of individual battery blocks. Sometimes it is possible to disassemble the battery and remove individual battery blocks while continuing to use other battery blocks. In some examples, a separate state prediction at the battery block level may be required.

[0029] The various examples described in this document can be implemented at least partially on the server side. This means that at least part of the logic associated with the state prediction can be executed on a central server separately from the battery or the device operated by the battery. For this purpose, a communication connection can be established in particular between the server and one or more battery management systems. By implementing at least part of the logic on the server, particularly accurate and computationally intensive models and / or simulations can be used in conjunction with the state prediction. This enables the state prediction to be performed particularly accurately. In addition, data from a group of batteries can be collected and used, for example in conjunction with a machine learning model.

[0030] The various examples described herein can perform a state prediction based on measurement data from the battery during use of the battery. This means that the state prediction is performed in particular at a specific point in time within the life cycle of the battery with a decrease in the SOH. The battery can then be in field use. In this way, in particular, the aging state of the battery up to now can also be taken into account. This also enables a particularly accurate predictive state determination.

[0031] Figure 1 Some aspects related to the system 80 are shown. The system 80 comprises a server 81 connected to a database 82. In addition, the system 80 comprises a communication connection 49 between the server 81 and each of the plurality of batteries 91-96. The communication connection 49 may be implemented, for example, via a mobile communication network.

[0032] Figure 1 By way of example, the battery 91-96 can send the status data 41 to the server 81 via the communication connection 49. For example, it is possible that the status data 41 indicates one or more operating values ​​of the respective battery 91-96. The status data 41 can be sent in an event-driven manner or according to a predetermined time scheme.

[0033] As a general rule, the operating value can, for example, indicate the actual value of the SOH of the corresponding battery 91-96. For example, the operating value can therefore indicate the capacity of the corresponding battery 91-96. The operating value can alternatively or additionally also indicate the actual value of the impedance of the corresponding battery 91-96. However, it is also possible that the operating value indicates the operating configuration of the corresponding battery 91-96. For example, an event-related operating configuration and / or a collective operating configuration can be indicated by the operating value.

[0034] Figure 1 It is also shown by way of an exemplary example that the server 81 can send control data 42 to the batteries 91 - 96 via the communication connection 49 . It is possible, for example, that the control data 42 can indicate one or more operating limits for the future operation of the respective battery 91 - 96 . It is conceivable, for example, that these operating limits are determined based on a human-machine display. For example, the control data can indicate one or more control parameters for the thermal management of the respective battery 91 - 96 (i.e., for example, the selection of the cooling capacity) and / or the charging management of the respective battery 91 - 96 (i.e., for example, the limits of the charging current). The server 81 can thus influence or control the operation of the batteries 91 - 96 by using the control data 42 .

[0035] In addition, Figure 1 91 - 96 , a corresponding SOH 99 is also schematically shown. As a general rule, the SOH 99 of a battery 91 - 96 may include one or more different characteristic parameters depending on the implementation. Typical characteristic parameters of the SOH 99 may be, for example: capacitance, that is, the maximum possible stored charge; and / or electrical impedance, that is, the frequency response of the resistance or AC resistance as a ratio of voltage to current intensity.

[0036] Next, a state prediction technique is described which enables the SOH 99 of each battery 91-96 to be predicted within a prediction interval during the use of the battery 91-96. This means that, for example, the electrical impedance and / or the capacity can be predicted.

[0037] Figure 2 Some aspects related to the batteries 91-96 are shown. The batteries 91-96 are connected to a corresponding device 69. This device is driven by the power from the corresponding battery 91-96.

[0038] The batteries 91-96 include or are associated with one or more management systems 61, such as a BMS or other control logic such as an on-board unit in the case of a vehicle. The management system 61 can be run on a CPU, for example, by software. Alternatively or additionally, an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA) can be used. The batteries 91-96 can communicate with the management system 61, for example, via a bus system. The batteries 91-96 also include a communication interface 62. The management system 61 can establish a communication connection 49 with the server 81 via the communication interface 62.

[0039] exist Figure 2 The management system 61 is shown separately from the batteries 91 - 96 , but in other examples it is also possible that the management system 61 is part of the batteries 91 - 96 .

[0040] Furthermore, the batteries 91-96 include one or more battery blocks 63. Each battery block 63 typically includes a number of battery cells connected in parallel and / or in series. Electrical energy can be stored there.

[0041] Typically, the management system 61 may rely on one or more sensors in one or more battery blocks 63. For example, the sensor may measure the current and / or voltage in at least some of the battery cells. The sensor may alternatively or additionally also measure other parameters in conjunction with at least some of the battery cells, such as temperature, volume, pressure, etc. The management system 61 may then be configured to determine the current SOH99 of the corresponding battery 91-96, or optionally determine the current SOH99 of each battery block based on one or more such measured values ​​from the sensor, that is, for example, determine the capacity and / or electrical impedance of the battery and send it to the server 81 in the form of operating data. The management system 61 may also be configured to implement thermal management and / or charging management of the corresponding battery 91-96. In conjunction with thermal management, the management system 61 may, for example, control cooling and / or heating. In conjunction with charging management, the management system 61 may, for example, control the charging rate or the depth of discharge. The management system 61 may therefore set one or more operating boundary conditions for the corresponding battery 91-96.

[0042] Figure 3 Some aspects related to the server 81 are shown. The server 81 includes a processor 51 and a memory 52. ​​The memory 52 may include a volatile storage element and / or a non-volatile storage element. In addition, the server 81 may also include a communication interface 53. The processor 51 may establish a communication connection 49 with each battery 91-96 and the database 82 via the communication interface 53. The server 81 may also have an interface to communicate with a user, for example, also with a user of one of the batteries 91-96. This means that the server 81 can be configured for human-computer interaction.

[0043] The program code may be stored in the memory 52 and loaded by the processor 51, for example. The processor 51 may then execute the program code. The execution of the program code causes the processor 51 to implement one or more of the following processes as described in detail in conjunction with the various examples herein: characterizing the battery 91-96; performing one or more state predictions of one or more batteries 91-96, for example based on operating values ​​received as state data 40 from the corresponding battery 91-96 via the communication link; performing electrical simulation of the battery 91-96; performing thermal simulation of the battery 91-96; performing aging assessment of the battery based on one or more operating configurations; sending control data to the battery 91-96, for example, in order to set operating boundary conditions; storing the characterization results of the corresponding battery 91-96 in the database 82; performing human-computer interaction based on the characterization results of the corresponding battery 91-96, for example, to select operating limits of the operating configuration, etc.

[0044] Figure 4 is a flow chart of an exemplary method. The method is implemented by a server. The method is used to characterize a battery on the server side. For example, it is possible that according to Figure 4 The method is implemented by the processor 51 of the server 81 based on the program code from the memory 52 (see Figure 3 ).exist Figure 4 Optional blocks are indicated by dashed lines.

[0045] First, one or more operating values ​​of the battery to be characterized are obtained in block 1001. For this purpose, status data are received in block 1001, for example, via a communication connection between the battery and a server.

[0046] The one or more operating values ​​may, for example, relate to the SOH of the battery. The one or more operating values ​​may, for example, relate to the capacity of the battery and / or the impedance of the battery. It is generally also possible to indicate one or more additional or other characteristic parameters of the operation of the battery by one or more operating values. For example, in some examples, current data (e.g., time series) and / or voltage data (e.g., time series) may be indicated by the operating values. This means that the operating value may, for example, describe the time variation of the current in one or more cells of a battery block of the battery or may describe the time variation of the voltage in one or more cells of a battery block of the battery. For example, the operating value may also describe the temperature in one or more regions of the battery. The operating value may, for example, describe the corresponding time series of the temperature data. Alternatively or additionally, it is also possible that the operating value indicates one or more operating configurations of the battery within a monitoring interval.

[0047] As a general rule, an operating profile may, for example, describe one or more of the following characteristic parameters for the operation of the battery: the used depth of discharge; the average depth of discharge; the used charge rate; the used discharge rate; the typical temperature during operation; the frequency of the charging process; the frequency of the discharging process; the duration between two charging processes; the duration between two discharging processes; etc. Event-related operating profiles and / or collective operating profiles may be used.

[0048] Next, in block 1002, one or more state predictions of the battery are performed to characterize the battery.

[0049] If a plurality of state predictions are performed in block 1002, these state predictions can be associated with different boundary conditions for the operation of the battery. The operating boundary conditions considered in block 1002 can relate, for example, to one or more of the following elements: control parameters for the thermal management of the battery and / or control parameters for the charge management of the battery. The operating boundary conditions can generally determine specific conditions for the operation of the battery, which are separated from a specific operating configuration, which is determined, for example, by using a corresponding device 69 associated with the corresponding battery (that is, for example, load, charge drawn, discharge rate, charge rate, depth of discharge, etc.).

[0050] In block 1002, the one or more state predictions may be based on an operating configuration derived from an operating configuration indicated by operating values ​​from block 1001 for the corresponding battery during a monitoring interval. It is possible, for example, to determine the operating configuration for the one or more state predictions in block 1002 based on measurements of the battery during the monitoring interval. This means that, for example, a measured DOD and / or a measured SOC cycle and / or a measured charge rate etc. may be used in conjunction with the one or more state predictions in block 1002. By using an operating configuration for the one or more state predictions in block 1002, a particularly reliable or accurate state prediction may be achieved, wherein the operating configuration is oriented to a specific operation of the corresponding battery within the monitoring interval.

[0051] The state prediction can result in a time curve of the aging within the prediction interval. Figure 5 Shown in.

[0052] Figure 5 Shown with batteries, for example Figure 1 Some aspects related to the aging of one of the batteries 91-96. Figure 5151 shows the SOH 99 as a function of time. During the monitoring interval 151, the SOH 99 decreases. This decrease in the SOH 99 can be measured, for example, by a corresponding management system (see Figure 2 , where it is shown that the management system 61) uses operating values ​​for the capacity and impedance of the respective battery.

[0053] The battery is then characterized by performing a plurality of state predictions 181-183 for the battery at a point in time 155. The state predictions 181-183 provide a prediction of the battery aging as a result, that is to say, the SOH99 during the prediction interval 152. Figure 5 It can be seen that SOH 99 varies between different state predictions 181-183 due to the different operating boundary conditions used. The operating boundary conditions can differ in temperature during charging and discharging, open circuit charge state, charge rate and discharge rate, charge end point and discharge end point, cycle depth and / or average charge state, and combinations thereof. For example, it is possible that the state prediction 181 generated as a function of time during the prediction interval 152 when SOH 99 drops slightly adopts other configurations of thermal management compared to the state prediction 183 generated as a function of time during the prediction interval 152 when SOH 99 drops more strongly. Thermal management of the battery can, for example, achieve a lower operating temperature as an operating boundary condition for the state prediction 181 through active cooling.

[0054] In the previous example, in which the state predictions 181-183 are associated with different operating boundary conditions, it is also possible to change the operating configuration for multiple state predictions as an alternative or in addition in other examples. For example, different state predictions can be associated with different DODs and / or different maximum SOCs and / or different charging rates, etc. This means that the battery operation strategy can be changed.

[0055] Now refer to Figure 4 : As a general rule, the results of one or more of the state predictions 181-183 from block 1002 can be evaluated in different ways.

[0056] In one example, it is possible to control a management system associated with the corresponding battery based on the results of the one or more state predictions from block 1002, see block 1003. In this interrelationship, it is possible, for example, to perform a plurality of state predictions in block 1002 and to compare the corresponding results of the plurality of state predictions with each other. The management system can then be controlled based on the comparison of the results. This can be done based on the results, for example, taking into account human-computer interaction. For example, it is possible to determine control data based on the comparison of the results (see Figure 1: control data 42) and send these control data to the management system. As a general rule, one or more different parameters for the operation of the corresponding battery can be set. For example, it is possible that the control data specifies one or more operating limits for the future operation of the battery. Alternatively or additionally, it is also possible that the control data specifies one or more control parameters for the thermal management and / or charging management of the battery. By means of such feedback of the results of one or more state predictions or generally of the characterization of the battery, a particularly long-lasting operation of the corresponding battery can be achieved.

[0057] Some examples are further explained below: For example, the operating limits of the future operation of the corresponding battery can be set so that the battery ages only slowly, that is, the SOH 99 only decreases slowly over time. For example, it is conceivable that the results of the characterization are output to the user of the corresponding battery through a suitable human-machine interface. This can be done, for example, via a website, where state predictions for different operating configurations of the battery can be shown. Mobile apps can also be considered. When the results of the characterization are available, a push message can be sent. The user can then select one of the battery configurations and can then adjust the future operation of the corresponding battery accordingly, so as to achieve the selected operating configuration with the help of operating boundary conditions. The user can therefore make a decision on how the battery should be charged in the future, taking into account the expected aging and taking into account the limitations on the battery operation caused by selecting a specific operating configuration.

[0058] For example, if the state prediction identifies specific operating boundary conditions of the battery caused by particularly strong aging of the battery, these operating boundary conditions can be excluded from the future operation of the battery by appropriately defining the operating limits. In this way, it is possible, for example, to compensate for manufacturing errors of the battery by appropriately controlling the operation. For example, it is possible that different state predictions are associated with different operating boundary conditions, that is, for example, with other operating parameters for thermal management or charging management (such as charging current). Then, it can be inferred from the aging comparison of the selections for different operating boundary conditions how the operating limits should be selected in combination with the operating boundary conditions to avoid excessive aging. This selection can be made, for example, by a user, for example, through a front-end call to a server or other human-machine interface. The user can therefore, for example, determine the maximum charging current, and the impact of this selection on aging can be fed back to the user based on the representation described here.

[0059] In another example, if a specific operating configuration of a battery caused by particularly strong aging of the battery is identified within the scope of the state prediction, these operating boundary conditions can be excluded from the future operation of the battery by appropriately defining the operating limits. For example, it is possible to associate different state predictions with different operating configurations. Then the aging can be compared with the selection of different operating boundary conditions, and it can be determined how the operating limits associated with the operating configuration should be selected to avoid excessive aging. A specific example is described below for this purpose: For example, a microgrid with a static energy storage device can be set as a solar system that can be repeatedly charged. The microgrid can have a connection to the main grid and operate the so-called peak regulation, that is, by feeding energy into the main grid to mitigate the load peak of the main grid (usually at night). However, the battery of the static energy storage device is charged at noon because the energy generated by the solar system is particularly strong at that time. The operating configuration can refer to the charging rate and discharge rate of the battery. Then the operating configuration can be adjusted according to the state prediction. That is to say, different operating strategies can be determined for the microgrid. In another example, a state prediction for different charging strategies can be performed for battery-operated products in mobile applications, such as electric vehicles or electric buses. Based on this, the charge management and / or thermal management of the battery can be adjusted before or during charging to prevent excessive aging.

[0060] As a general rule, one or more predetermined use ranges for the future operation of the battery can optionally also be taken into account when determining the control data. This means that, although one or more specific operating ranges are identified as being particularly harmful to the aging of the battery in conjunction with the state prediction, the operation of the battery cannot be restricted to the corresponding operating ranges based on the predetermined preferences. These use ranges may influence the operating boundary conditions and / or the operating configuration. This can ensure that, for example, no functional restrictions are caused depending on the type of battery-powered device.

[0061] However, it is not necessary in all examples to feed the results of the one or more state predictions back into the operation of the battery. In this regard, block 1003 is an optional block.

[0062] In some examples, it is also possible, alternatively or additionally, to store the results of the state prediction in a database (see Figure 1 : database 82), see block 1004. This storage will enable a particularly rapid determination of the state of aging of the battery, for example, by subsequently comparing the actual operating configuration during the prediction interval 152 with the stored results of one or more state predictions. For example, a secondary use of the battery can be selected in a targeted manner depending on whether the SOH 99 still enables a corresponding use of the battery. In addition, a machine learning model can be trained for aging assessment.

[0063] In some examples, it is alternatively or additionally possible to output the result of the state prediction to a user, for example via a suitable human-machine interface. In this way, the user can, for example, appropriately intervene in the operation of the battery to avoid excessive aging in view of the state prediction. The operating boundary conditions can be set by the user.

[0064] Executing one or more state predictions in block 1002 may require more computing resources. This means that the corresponding server, which is typically tasked with characterizing multiple batteries, can selectively trigger the execution of one or more state predictions based on one or more trigger criteria (at time point 155, see Figure 5 ): This avoids the need to constantly perform new state predictions. More precisely, the one or more state predictions can be performed in a demand-oriented manner. For example, it is possible that the server repeatedly obtains the (e.g. collective or event-related) operating configuration of the battery within the monitoring interval 151 from a corresponding management system associated with the battery. The server can then monitor the development of the operating configuration over time. This may mean that the server will check changes in the operating configuration between two points in time. Based on this monitoring, the execution of at least one state prediction can then be selectively triggered. If, for example, one or more characteristics of the operating configuration change, the execution of the state prediction can be triggered: based on the recognition that in this situation there may be a significant impact on the aging state of the battery, particularly due to changes in the state configuration.

[0065] However, in other simple implementations, it is also possible to execute one or more state predictions in block 1002 according to a predetermined schedule, for example, at a specific repetition rate.

[0066] Next, combine Figure 6 The flowchart in FIG. 100 illustrates an exemplary implementation of one or more state predictions described in block 1002 .

[0067] Figure 6 is a flow chart of an exemplary method. Figure 6 The method may be implemented by the server. For example, it is possible that according to Figure 6 The method is implemented by the processor 51 of the server 81 based on the program code from the memory 52 (see Figure 3 ).

[0068] according to Figure 6 The method is used to predict the state of battery health. If multiple state predictions are to be performed, multiple executions are performed according to Figure 6 method.

[0069] In block 1011, firstly operating values ​​of capacity and impedance for the respective battery are obtained. Block 1011 therefore corresponds to block 1001. This means that the current value of the SOH 99 of the battery is obtained. This is typically performed based on state data received from the respective management system associated with the respective battery. These operating values ​​are used to initialize the state prediction.

[0070] A number of iterations 1099 of blocks 1012 - 1014 are then carried out. The different iterations 1099 correspond to time steps of the state prediction, that is to say the time that is advanced during the prediction interval 152 .

[0071] In block 1012 , for a corresponding time step of a corresponding iteration 1099 , a simulation of the electrical state of the battery and the thermal state of the battery is first performed using a corresponding simulation module.

[0072] In block 1012 , a simulation is performed taking into account the corresponding operating boundary conditions of the battery. This depends on the corresponding state predictions 181 - 183 . In addition, a corresponding operating configuration for the battery operation can be assumed.

[0073] The electrical simulation module may be coupled with the thermal simulation module to simulate electrical conditions and thermal conditions.

[0074] The electrical simulation module can use an equivalent circuit model (ECM) of the battery. The ECM can include electrical components (resistance, inductance, capacitance). The component parameters of the ECM can be determined, for example, with the aid of a Nyquist diagram of a characteristic frequency range of the transmission characteristics of the cell block with the battery. By means of an implementation on the server 81, the number of RC elements can be selected to be particularly large, for example more than three or four. This can achieve a particularly high accuracy of the electrical simulation. In this case, one ECM can be used for each cell in the cell block.

[0075] The thermal simulation model enables the determination of the temperature variation curve over time and the optional local temperature. Here, heat sources and heat sinks can be considered. Heat dissipation in the surrounding environment can be taken into account. Details of the heat generation model are described, for example, in D.Bernandi, E.Pawlikowski and J.Newman's "General Energy Balance of Battery Systems" (Journal of the Electrochemical Society, 1985). Analytical or numerical models for local temperature distribution can be used. The influence of the thermal management system can be taken into account. For example, M.-S.Wu, KHLiu, Y.-Y.Wang and C.-C.Wan's "Heat Dissipation Design of Lithium-Ion Batteries" (Energy Journal, Vol. 109, Reference No. 1, pp. 160-166, 2002).

[0076] Based on this, an aging assessment is then performed in block 1013 , that is to say the capacity and the impedance of the battery in the corresponding time step are determined based on the results of the simulation of the electrical and thermal state of the battery.

[0077] Different techniques can be used in conjunction with the aging assessment. The aging assessment can, for example, include an empirical aging model and / or a machine-learned aging model. For example, an empirical aging model and a machine-learned aging model can be used in parallel and the results of the two aging models can then be combined by averaging, for example, weighted averaging.

[0078] As a general rule, the empirical aging model may include one or more empirically determined parameters that relate the operating configuration of the battery obtained from the simulation of block 1012 to a degradation of the SOH 99, such as a decrease in capacity and / or an increase in impedance. The parameters may be determined, for example, in laboratory measurements. In J. Schmalstieg, S. An exemplary empirical aging model is described in M. Ecker and D. U. Sauer, "A Holistic Aging Model for 18650 Li-ion Batteries Based on Li(NiMnCo)O2," Journal of Energy, Vol. 257, pp. 325-334, 2014.

[0079] In contrast, machine-learning aging models can be continuously adapted by machine learning based on the state data obtained from different batteries of the same type. For example, artificial neural networks such as convolutional networks can be used. Other techniques include so-called support vector methods. For example, a set of batteries (see Figure 1 : The data of batteries 91-96) is used to train the corresponding algorithm through machine learning.

[0080] Different aging models can receive different inputs from the simulation in block 1012. In a variant, it is possible, for example, to determine one or more simulated, collective operating profiles (sometimes also referred to as load profiles) for the battery based on the simulated electrical states and / or simulated thermal states determined in multiple iterations 1099. Such collective operating profiles can respectively specify the frequency of operation of the battery at a specific operating point. This means that the collective load profile quantifies the frequency of occurrence of battery operation at a specific operating point similar to a histogram, but does not specify the operating sequence of the battery at the operating point or the specific time point of the battery operation at a specific operating point. Figure 7An exemplary, simulated, collective operating configuration 500 is shown. The figure illustrates the corresponding frequency 509 (in arbitrary units) of the battery operation for different operating points defined in two dimensions, namely depth of discharge (DOD) and state of charge (SOC). SOC illustrates the current charge as a percentage of the battery's capacity. However, the two dimensions DOD-SOC are only examples here. Alternative dimensions that can be observed in conjunction with the collective operating configuration for aging assessment include, for example: charging rate; temperature during charging; temperature during discharge; discharge rate; etc. It is also possible to set more than two dimensions for the collective operating configuration, for example, three or four dimensions.

[0081] The frequency can be indicated, for example, by calendar or cycle. Different counting methods can be used here, for example, peak value counting, over-limit counting, instantaneous value counting, range counting, range-to-average value counting, rain flow counting (see, for example, DE 10 2009 006 461 A1), etc.

[0082] For example, Figure 7 The darkening in FIG. 7 shows that a particularly strong aging of the battery occurs at a low state of charge for a large depth of discharge.

[0083] In another variant, as an alternative to or in addition to one or more simulated, collective operating configurations, one or more event-related operating configurations may also be taken into account in conjunction with one or more aging models used. Figure 8 51 is shown. It shows the time curves of the charging rate, the voltage between the battery terminals, and the temperature associated with the charging event of the corresponding battery. In addition, the corresponding average value 511 is also shown.

[0084] In some examples, it is conceivable that one or more parameters characterizing an event are considered directly in conjunction with the aging model (in Figure 8 time curve of the charge rate, voltage and temperature); however, it is also conceivable that the derived parameters (in Figure 8 In any case, the average value is 511. Figure 8 It can also be seen from the example that one or more parameters of the event-based operation configuration are not associated with the frequency of the corresponding operation point, but are related to individual events.

[0085] It is possible that the aging assessment includes a plurality of components, for example a first model component, which receives as input one or more simulated, collective operating configurations, and a second model component, which receives as input one or more simulated, event-related operating configurations. The aging assessment can then be designed particularly accurately by taking these different operating configurations into account. For example, in this interrelationship it is possible to firstly carry out the different components of the aging assessment individually and then combine the results of the different components of the aging assessment with one another in order to obtain the result of the aging assessment.

[0086] Reference again Figure 6 : Then, in block 1014, a check is performed to determine whether the interruption criterion is met. If not, block 1012 is executed again for the next time step within the prediction interval 152, that is, for the next iteration 1099. In this case, the capacity and impedance determined in the previous iteration 1099 are used, that is, the simulations in block 1012 are built upon each other. This iterative adjustment of the capacity and impedance enables a particularly accurate state prediction.

[0087] The state prediction is terminated when an interruption criterion is met in block 1014. Examples of interruption criteria include: number of iterations 1099; end of prediction interval 152; exceeding or falling below a capacity and / or impedance threshold; and the like.

[0088] In many examples it may be desirable to have the initial operating values ​​for impedance and capacity in block 1011 available with particularly high accuracy. Fig. 9 , describes an exemplary technique that achieves this.

[0089] Fig. 9 is a flow chart of an exemplary method. Fig. 9 The method can be performed, for example, on the device side on a management system associated with the battery. For example, it is possible that according to Fig. 9 The method is based on Figure 2 But it can also be executed on the server side Fig. 9 method.

[0090] according to Fig. 9 The method enables particularly accurate determination of capacity and impedance based on the measured data. The corresponding actual values ​​can be determined on site during normal operation.

[0091] First, in block 1021, an operating value for the impedance is determined. This can be done, for example, when using an ECM. For example, the operating value of the impedance can be determined by determining parameters of one or more components of the ECM. This means, for example, that the resistance of an R element of the ECM can be determined or the inductance of an L element or the capacitance of a C element can be determined. It is possible, for example, to determine the operating value based on a time series for the current and / or voltage.

[0092] Next, an operating value for the capacitance is determined in block 1022. This can be done, for example, based on the open circuit voltage, as described in M. Einhorn, F. V. Conte, C. ral and J. Fleig, "Method for online capacity estimation of lithium-ion battery cells using state of charge and transfer charge" (IEEE Transactions on Industrial Applications, Vol. 48, Ref. 2, pp. 736-741, 2012).

[0093] The operating values ​​determined in blocks 1021-1022 are then transmitted, for example, via a corresponding communication interface in the form of status data (see Figure 1 :State data 41) is transmitted to the server. In this regard, block 1023 corresponds to block 1001 or 1011.

[0094] Fig.10 Combined with, for example, Fig. 9 Block 1021 determines an operating value for the impedance of the battery to illustrate some aspects. Fig.10 The corresponding technology shown in the example of can be implemented, for example, on a local management system associated with the battery (for example, according to Figure 2 The result of the corresponding evaluation can then be transmitted to the server 81, for example, in the form of status data 41. This means that the operating value for the impedance can be determined on the device side at the battery. However, a corresponding determination can also be made on the server side.

[0095] Fig.10 A time series of measured values ​​of the current in one or more cells of the battery is shown. Based on the time profile of the current (or alternatively or additionally with the aid of the time profile of the voltage and / or other characteristic parameters of the battery's electrical operation), an operating value for the impedance of the battery can be determined. One or more characteristic variables of the thermal operation can also be taken into account.

[0096] According to various examples, this is achieved by Fig.10In the example of , several values ​​are extracted from the time series of current and / or voltage within the jump windows 711, 712. One or more analysis algorithms can then be applied to the extracted values ​​for each jump window. For example, one or more analysis algorithms can be used for the values ​​within the jump window 711 and one or more analysis algorithms can then also be used for the values ​​within the jump window 712. By extracting several values ​​from a continuous time series, real-time processing can be achieved by reducing the data basis. The one or more analysis algorithms can be operated sequentially and / or in parallel based on the values ​​extracted from specific jump windows 711, 712. Candidate values ​​for impedance can then be obtained for different jump windows 711, 712, and the candidate values ​​can then be combined to obtain corresponding operating values.

[0097] These techniques are based on the knowledge that it is often possible to extract a time profile of a current and / or voltage which is particularly relevant to the impedance from a continuous time series of the corresponding data.

[0098] In particular, it is possible to determine one or more jump windows 711, 712 in such a way that they include relevant data. One or more jump windows 711, 712 can generally be determined by analyzing the time series of current data and / or voltage data with respect to one or more criteria. For example, the criteria can be selected from the following group: duration 722 of the discharge operation; charge quantity of the discharge operation; rest phase 721 before the discharge operation; or also in combination with one or more such parameters of the charging process (instead of the discharge process).

[0099] These techniques are based on the recognition that in vehicle applications strongly fluctuating and undefined current excitations often occur in the field, instead of the defined relaxation times and pulse loads used in laboratory parameterization. In order to allow meaningful parameter values ​​to be determined for the ECM in conjunction with the determination of the impedance operating value, the above-mentioned jump windows 711, 712 are selected in the window search process. Here, the starting point is, for example, a circular buffer of a defined length. Here, the buffer stores the unfiltered current and voltage measurement values ​​of the BMS at the sampling rate according to the first-in-first-out principle (FiFo). In each time step, the entire circular buffer is compared with a set of predefined sample windows, which define the criteria. If a certain area meets the window of the sample criteria, then the measurement value is extracted for further analysis.

[0100] Examples of corresponding criteria include: For parameter determination, the initial value of the RC element must be known. If a constant current exists within a sufficiently long time period 721, then the corresponding resistance of the ECM can be determined. The start of the jump window 711, 712 should therefore be defined after a time interval 721 with a current that is as constant as possible. The second criterion for the selection of the jump window relates to the dynamics of the current and voltage signals within the jump window 711, 712 itself. It should be noted here that the dynamic effects to be mapped (such as diffusion) are also sufficiently stimulated by the current load. If, for example, a window with too small a charge flow is selected at the beginning of driving when the battery cell is relaxed, then the diffusion effect may appear too weak and the corresponding parameter determination may be unreasonable. Therefore, in addition to the current and voltage, for example, the incremental charge amount is also taken into account. Under effects with a significant current rate dependency, such as the passage of charge carriers, current jumps that are as step-like as possible should also be identified in order to be able to correctly align the current rate dependencies of the parameters.

[0101] Fig.11 Combined with, for example, Fig. 9 Block 1022 of determining an operating value for the capacity of the battery illustrates some aspects. Fig.11 The corresponding technology shown in the example of can be implemented, for example, on a local management system associated with the battery (for example, according to Figure 2 The result of the corresponding evaluation can then be transmitted to the server 81, for example, in the form of status data 41. This means that the operating value for the impedance can be determined on the device side at the battery. However, a corresponding determination can also be made on the server side.

[0102] Fig.11 The stable voltage (open circuit voltage, OCV) is shown as a function of the state of charge (ampere-hours). The state of charge is related to the SOC. OCV is related to the state of charge through the open circuit voltage characteristic curve. Fig.11 7 shows two open circuit voltages 751, 752 and the associated charge states 761, 762. The charge states 761, 762 are determined taking into account a predetermined characteristic curve 770. For example, a table lookup can be implemented. From a linear interpolation of the two charge states 761, 762 (taking into account the characteristic curves 770-772, which may be compressed if necessary) in conjunction with the associated charge flow, the capacity of the battery can be inferred.

[0103] In various examples described herein, a compression factor can be determined for characteristic curve 770 in order to obtain compressed characteristic curves 771, 772. The compression factor is determined based on an additional aging evaluation of the battery that is different from an aging evaluation that is performed immediately based on a specific capacity (that is, the additional aging evaluation is different from, for example, the aging evaluation performed based on the specific capacity). Figure 6 The further aging evaluation and the aging evaluation performed based on the determined capacity can be independent of each other. This is based on the recognition that the characteristic curve 770 can change with the aging state.

[0104] In general, different techniques are usually possible in combination with the further aging assessment. In one example, it is conceivable, for example, to monitor the lithium insertion potential in combination with a lithium-ion battery. The lithium insertion potential is, for example, a peak in a dV / dQ curve during the charging process of a battery. They represent the transition between different phases of the lithium insertion process in the graphite anode. Their height and / or their spacing characterize the aging state of the battery and can therefore be used within the scope of the method.

[0105] The features of the above-described embodiments of the present invention can of course be combined with one another. The features described are intended to be used not only in the described combination, but also in other combinations or alone, without departing from the scope of the present invention.

[0106] For example, different techniques related to battery state prediction may be described. These techniques may also be applied to individual blocks of batteries or cells or groups of cells of batteries.

[0107] In addition, techniques related to characterization and / or state prediction of the battery are described. These results can be used to set the future operating configuration of the battery. It is contemplated that these results can be output to the user of the battery through a suitable human-machine interface. The user is then enabled to select a future operating configuration based on this information about the state of the battery. This corresponds to a human-machine interaction for controlling the battery.

Claims

1. A method for characterizing a rechargeable battery (91-96) on a server side, the method comprising: - obtaining operating values ​​for the capacity of the battery (91-96) and the impedance of the battery (91-96), - based on the operating values: performing at least one state prediction (181-183) of the battery (91-96), wherein each state prediction of at least one state prediction (181-183) comprises a plurality of iterations (1099), wherein in each iteration (1099) a simulation of the electrical state of the battery (91-96) and of the thermal state of the battery (91-96) is performed and based on the results an aging assessment for capacity and impedance is determined, wherein the aging assessment from a first iteration (1099) of the corresponding state prediction (181-183) is used for the simulation in a subsequent second iteration (1099) of the corresponding state prediction (181-183), wherein each state prediction of the at least one state prediction (181-183) is associated with a corresponding operating boundary condition of the battery (91-96), wherein the at least one state prediction (181-183) comprises a plurality of state predictions (181-183) associated with different operating boundary conditions, Wherein, the method further comprises: - comparing the results of the plurality of state predictions (181-183), and - Based on the comparison of the results: determining control data (42) for a management system (61) associated with said battery (91-96) and sending the control data (42) to the management system (61).

2. The method according to claim 1, in, The control data (42) indicates one or more elements of the following group: one or more operating limits (552) for future operation of the battery (91-96), wherein the operating limits (552) are associated with one or more operating boundary conditions or operating configurations of the battery (91-96); control parameters for thermal management of the battery (91-96); and control parameters for charging management of the battery (91-96).

3. The method according to claim 2, in, The control data (42) are also determined based on a predetermined range of use (551) for future operation.

4. The method according to any one of claims 1 to 3, in, The operating boundary conditions involve one or more of the following elements: control parameters for thermal management of the battery (91-96); and control parameters for charge management of the battery (91-96).

5. The method according to any one of claims 1 to 3, in, Each state prediction of the at least one state prediction (181-183) is associated with a corresponding operating configuration of the battery (91-96).

6. The method according to any one of claims 1 to 3, further comprising: - based on the electrical states simulated and / or the thermal states simulated in a plurality of iterations (1099) of the at least one state prediction (181-183): determining a collective operating configuration (500) for one or more simulations of the battery (91-96), the operating configurations respectively specifying the frequency of operation of the battery (91-96) at a specific operating point, and - optionally using the collective running configuration (500) of said one or more simulations for aging assessment in said iteration (1099).

7. The method according to any one of claims 1 to 3, further comprising: - based on the simulated electrical state and / or the simulated thermal state in at least one iteration (1099) of the at least one state prediction (181-183): determining one or more simulated event-related operating configurations (510) for the battery (91-96), and - optionally using one or more simulated, event-related run configurations (510) for aging evaluation in said iteration (1099).

8. The method according to claim 6, in, The aging assessment in the iteration (1099) includes a first component based on a collective operating configuration of the one or more simulations, wherein the aging assessment in the iteration (1099) includes a second component of the event-related operating configuration based on the one or more simulations, Wherein, the method further comprises: - combining the results of the first component of the aging assessment and the results of the second component of the aging assessment.

9. The method according to any one of claims 1 to 3, in, The aging assessment in the iteration (1099) uses at least one of an empirical aging model and a machine-learned aging model.

10. The method according to any one of claims 1 to 3, further comprising: -determine the operating value for the impedance, Among other things, determining the operating value for impedance includes: - extracting values ​​for one or more hopping windows (711, 712) from the time series of current-voltage data of the battery (91-96), - for each hopping window (711, 712): applying an analytical algorithm to the extracted values ​​to determine a corresponding candidate value for the impedance using an equivalent electrical circuit model, and - Combining candidate values ​​associated with a plurality of hopping windows (711, 712) to obtain a running value.

11. The method according to claim 10, further comprising: - determining the one or more hopping windows (711, 712) based on an analysis of a time series of current-voltage data with respect to one or more standards (721, 722), The one or more criteria are selected from: the duration of the discharge process; the charge amount of the discharge process; the static phase (721) before the discharge process; the duration of the charging process; and the charge amount of the charging process.

12. The method according to any one of claims 1 to 3, further comprising: - determine the operating value for capacity, Among them, determining the operating value for capacity includes: - determining a first open circuit voltage in a first partial discharge state and a second open circuit voltage in a second partial discharge state, - determining a charge flow between a first partially discharged state and a second partially discharged state, - based on a predetermined characteristic curve: determining the state of charge (761) of the battery (91-96) for the first open circuit voltage, - based on a predetermined characteristic curve: determining a second state of charge (762) of the battery (91-96) for the first open circuit voltage, - determining a capacity based on the first state of charge, the second state of charge and the charge flow rate, and - Determining a compression factor of the characteristic curve based on another aging evaluation of the battery (91-96).

13. The method according to claim 12, further comprising: - performing said further aging assessment based on monitoring the lithium insertion potential of said battery (91-96).

14. The method according to any one of claims 1 to 3, further comprising: - obtaining the operating configuration of the battery (91-96) during the monitoring interval (151) from a management system (61) associated with the battery (91-96), - monitor the evolution of said running configuration, and - triggering the execution of said at least one state prediction (181-183) based on monitoring of said development.

15. The method according to any one of claims 1 to 3, further comprising: - obtaining the operating configuration of the battery (91-96) during the monitoring interval (151) from a management system (61) associated with the battery (91-96), Wherein, the at least one state prediction (181-183) is based on an operating configuration of the battery (91-96).

Images