Method for determining at least one calibration parameter of a system component of a cell system, cell system and computer program

Abstract

The invention relates to a method for determining at least one calibration parameter of at least one system component (30, 32), in particular a control unit (26), of a cell system (22) having at least one electrochemical cell (24). According to the invention, the calibration parameters are determined on the basis of at least one model of the cell system (22) from at least one system component characteristic parameter of the cell system (22) and from at least one system behavior characteristic parameter of the cell system (22).

Description

Technical Field

[0001] A method has been proposed for determining at least one calibration parameter of a system component of a cell system having at least one electrochemical cell. Summary of the invention

[0002] The invention is based on a method for determining at least one calibration parameter of at least one system component of a cell system, in particular a control unit, which cell system has at least one electrochemical cell.

[0003] It is proposed that the calibration parameter is determined based on at least one model of the cell system as a function of at least one system component characteristic parameter of the cell system and as a function of at least one system behavior characteristic parameter of the cell system.

[0004] The design of the method according to the invention makes it possible to achieve particularly precise parameterization of a cell system, preferably at least one system component of a cell system. Advantageously, the parameterization of the cell system can be achieved with particularly low sensor technology expenditure. Advantageously, a particularly effective regulation of the cell system can be achieved. Advantageously, a particularly effective operation of the cell system can be achieved.

[0005] The electrochemical cell is preferably configured as a solid oxide fuel cell. However, alternatively, it is also conceivable to configure the electrochemical cell as other fuel cells, for example, as a polymer electrolyte fuel cell, etc., as an electrolyte cell or as a storage cell. For example, it is also conceivable to use the electrochemical cell in an electrolyzer as an alternative. It is conceivable that the cell system has only one electrochemical cell, in particular the electrochemical cell already mentioned above, or that the cell system has a plurality of electrochemical cells, for example two, three or several hundred electrochemical cells. Preferably, the electrochemical cells of the cell system having a plurality of electrochemical cells are combined to form a stack and are in particular connected in series. It is also conceivable that the cell system comprises a plurality of stacks formed by a plurality of electrochemical cells, which are preferably combined into modules or towers.

[0006] The battery system has in particular a control unit. The control unit is in particular arranged to execute a method for determining at least one calibration parameter of at least one system component of the battery system. The control unit is in particular arranged to determine at least one calibration parameter based on at least one model according to at least one system component characteristic parameter of the battery system and according to at least one system behavior characteristic parameter of the battery system. "Setting" should be understood as specially configured, specially designed and / or specially equipped. The function that an object is set to determine should be understood as that the object implements and / or performs the specific function in at least one application state and / or operating state. The control unit in particular includes at least one processor and a storage element and an operating program stored on the storage element. The storage element is preferably configured as a digital storage medium, such as a hard disk or the like. The battery system in particular includes a housing, in which at least one electrochemical cell or a stack formed by a plurality of electrochemical cells is preferably arranged. It is conceivable that the control unit is at least partially arranged on the housing, in particular at least partially arranged in the housing. Alternatively or additionally, it is also conceivable that the battery system has an external unit, wherein the external unit includes at least a part of the control unit. The external unit can be, for example, a cloud, a server, and the like. It is conceivable that the external unit determines at least one calibration parameter based on at least one model from at least one system component characteristic parameter of the cell system and from at least one system behavior characteristic parameter of the cell system. It is also conceivable that the external unit comprises a storage element of the control unit.

[0007] The system component of the cell system is preferably a control unit. However, alternatively, it is also conceivable to configure the system component of the cell system as a fan, in particular an air fan or a circulation fan, a mass flow regulator, a heat exchanger, a flow sensor, a soft sensor or other system components of the cell system that are considered meaningful by a person skilled in the art. At least one system component characteristic parameter of the cell system can be, for example, a flow characteristic parameter of a fan (in particular a flow characteristic parameter of an air fan or a flow characteristic parameter of a circulation fan), a valve position of a valve (in particular a valve position of a needle valve), a pressure characteristic parameter of a fan (in particular a characteristic parameter related to a pressure buildup at the fan or a characteristic parameter related to a pressure reduction at the fan), a pressure characteristic parameter of a valve (in particular a characteristic parameter related to a pressure buildup at the valve or a characteristic parameter related to a pressure reduction at the valve), etc. At least one calibration parameter is provided for example to calibrate a flow sensor of air or fuel gas, calibrate an air flow control, calibrate a fuel gas flow control, calibrate a temperature regulation, calibrate a power regulation, calibrate a fan control, calibrate a soft sensor, etc. For example, the cell system comprises only at least one system component. However, alternatively, it is conceivable that the battery system has a plurality of system components, wherein at least one calibration parameter of at least one system component of the plurality of system components of the battery system is determined, preferably based on at least one model, from at least one, in particular previously mentioned, system component characteristic parameter of the battery system and at least one, in particular previously mentioned, system behavior characteristic parameter of the battery system, preferably at least one calibration parameter of each of the plurality of system components is determined. It is conceivable that at least one system component of the battery system has a system component characteristic parameter of the battery system, from which the calibration parameter of the at least one system component of the battery system is determined, in particular based on the model together with at least one system behavior characteristic parameter of the battery system. Alternatively, it is also conceivable that another system component of the plurality of system components of the battery system has at least one system component characteristic parameter of the battery system, from which the calibration parameter of the at least one system component of the battery system is determined, in particular based on the model together with the system behavior characteristic parameter of the battery system.

[0008] The characteristic parameter of the system behavior of the cell system has in particular at least one output characteristic parameter of the cell system and / or at least one operating parameter of the cell system. Preferably, the output characteristic parameter of the cell system is the output voltage of the cell system, in particular the output voltage of at least one electrochemical cell or a stack formed by a plurality of electrochemical cells, electrical work, electrical power, etc. The at least one operating parameter of the cell system is, for example, a requested load current, a temperature (in particular a gas temperature), a volume flow, an air utilization rate, a fuel utilization rate, a gas composition, a pressure, a pressure difference, etc. It is conceivable that the cell system, in particular the control unit or the external unit has a component detection unit, which is in particular configured to detect (preferably in the form of a sensor) at least one component characteristic parameter of the cell system. Alternatively or additionally, it is conceivable that the cell system (preferably a system component, in particular the control unit) or the external unit has an input unit for inputting at least one component characteristic parameter of the cell system. Preferably, the input unit is connected or can be connected to the control unit in terms of data technology, in particular for wireless and / or wired data transmission. The input unit can be configured, for example, as a keyboard, a keypad, a touch screen, a knob, a slide regulator, a button, etc.

[0009] The model can be a deterministic model (preferably a physical model) or a data-based model. The deterministic model, in particular the physical model, is preferably a simulation model of the battery cell system, which preferably describes the operation of the battery cell system in a steady state. Preferably, the deterministic model, in particular the physical model is based on the finite element method. The deterministic model (preferably the physical model) in particular has a plurality of input variables and at least one output variable. Preferably, the input variables of the deterministic model, in particular the physical model, include at least one system component characteristic parameter of the battery cell system. Preferably, the input variables of the deterministic model, in particular the physical model, include at least one calibration parameter. Preferably, the input variables of the deterministic model, in particular the physical model, include at least one operating parameter of the battery cell system. Preferably, at least one output variable of the physical model includes at least one output characteristic parameter of the battery cell system.

[0010] A deterministic model, in particular a physical model, preferably represents a mapping that maps input variables of the physical model (preferably at least one calibration parameter, at least one operating parameter of the battery cell system and at least one system component characteristic parameter of the battery cell system) into output variables of the physical model (preferably at least one output characteristic parameter of the battery cell system).

[0011] It is further proposed that, in order to determine the calibration parameters, the model, in particular the deterministic model (preferably the physical model), is inverted and optimized, preferably with respect to at least one calibration parameter. Preferably, in a method step, in particular in a data generation step, at least one data set for the deterministic model is generated, which has data on at least one system behavior characteristic parameter of the battery cell system, preferably data on at least one output characteristic parameter of the battery cell system and data on at least one operating parameter of the battery cell system. For example, the data set of the deterministic model (in particular the physical model) is generated by means of a statistical experimental design, for example by means of Latin hypercube sampling, active learning, etc. The data set of the deterministic model (in particular the physical model) preferably describes the target behavior of the battery cell system. Preferably, the data on at least one system behavior characteristic parameter of the battery cell system are generated from the data set of the deterministic model (in particular the physical model), so that the associated system state of the battery cell system can be described by the deterministic model, in particular the physical model. For example, it is conceivable that the deterministic model, in particular the physical model, can be applied only to stationary conditions, so that correspondingly sufficiently stationary system behavior characteristic parameters should be used. It is also conceivable to at least partially pre-process at least one system behavior characteristic parameter of the cell system, preferably at least one output characteristic parameter of the cell system, and / or at least one operating parameter of the cell system, for example by removing oscillations by means of filters or the like. It is conceivable to at least partially combine a plurality of system behavior characteristic parameters of the cell system. It is also conceivable to at least partially convert at least one system behavior characteristic parameter of the cell system, preferably by means of a control unit, to adapt it to a deterministic model, in particular a physical model, preferably in a data processing step. For example, if necessary, at least one system behavior characteristic parameter of the cell system from a data set of a deterministic model, preferably a physical model, is converted into an input variable and / or an output variable of the deterministic model, in particular a physical model, in particular in a data processing step. For example, it is conceivable that the cell system has a plurality of stacks, each stack having a plurality of electrochemical cells, wherein the individual stacks are averaged, or the detected current of the cell system is converted into a current density. Based on the deterministic model, in particular the physical model, at least one calibration parameter for at least one system behavior characteristic parameter of the cell system is preferably determined from the data set of the deterministic model, in particular the physical model, and at least one system component characteristic parameter of the cell system, in particular in a data processing step. For example, based on a deterministic model, in particular a physical model, preferably in a data processing step, at least one calibration parameter is determined, for which the deterministic model (in particular a physical model) together with at least one operating parameter of the battery cell system and at least one system component characteristic parameter of the battery cell system from a data set of the deterministic model (in particular the physical model) is used to predict at least one output characteristic parameter of the battery cell system from the data set of the deterministic model (in particular the physical model).In particular, it is preferred to apply an optimization method in the data processing step to determine at least one calibration parameter, such as Bayesian optimization, simulated annealing, gradient methods, etc. Advantageously, the at least one calibration parameter can be determined particularly accurately. Advantageously, a particularly effective system regulation of the cell system can be achieved.

[0012] Furthermore, it is proposed that the model is a deterministic model, in particular the above-mentioned deterministic model, which is used to train a regression model. Preferably, the regression model is trained based on results determined with the aid of a physical model. Preferably, the trained regression model is arranged to replace the deterministic model, in particular the physical model, at least for determining at least one calibration parameter based on at least one system behavior characteristic parameter of the battery cell system and at least one system component characteristic parameter of the battery cell system. The regression model is used in particular in an optimization method for determining at least one calibration parameter or for directly determining at least one calibration parameter based on at least one system behavior characteristic parameter of the battery cell system and at least one system component characteristic parameter of the battery cell system. The training of the regression model is based, for example, on linear regression, random forests, Gaussian processes, neural networks, interpretable boosting machines, etc. Advantageously, at least one calibration parameter can be determined in a particularly time- and computationally efficient manner with the aid of a regression model.

[0013] It is further proposed that as data on at least one system characteristic parameter of the cell system, data on at least one system characteristic parameter of at least one other cell system is at least partially used, the at least one other cell system having at least one electrochemical cell. The at least one system behavior characteristic parameter of the other cell system is in particular a variable similar to the at least one system behavior characteristic parameter of the cell system. Preferably, in a method step, in particular in a data provision step, a data set based on a data model is provided, preferably by means of a control unit. The data set based on the data model preferably has data on at least one system behavior characteristic parameter of the other cell system, preferably data on at least one output characteristic parameter of the other cell system and data on at least one operating parameter of the other cell system, which data are preferably detected at the other cell system during operation of the other cell system. The data set based on the data model has data on at least one system component characteristic parameter of a system component of the other cell system. The at least one operating parameter of the other cell system is in particular a variable similar to the at least one operating parameter of the cell system. The at least one output characteristic parameter of the other cell system is in particular a variable similar to the at least one output characteristic parameter of the cell system. Preferably, the other cell system has a detection unit for at least partially detecting data on at least one system behavior characteristic parameter of the other cell system at the other cell system. The detection unit is preferably connected to a control unit of the battery system in terms of data technology, in particular to an external unit of the battery system, preferably for wireless and / or wired data transmission. It is conceivable that data about at least one output characteristic parameter of the other battery system and / or data about at least one operating parameter of the other battery system are detected at the other battery system, preferably in a sensor manner, preferably with the aid of the detection unit. The detection unit comprises, for example, at least one sensor unit for at least partially detecting data about at least one system behavior characteristic parameter of the other battery system. The detection unit is preferably configured to detect at least data about at least one output characteristic parameter of the other battery system at the other battery system, in particular in a sensor manner. The sensor unit comprises, for example, at least one voltage sensor, in particular for detecting data about an output characteristic parameter of the other battery system, which is configured as an output voltage, at the other battery system. It is conceivable that the detection unit is configured to at least partially detect data about at least one operating parameter of the other battery system at the other battery system. Additionally or alternatively, it is conceivable that data about at least one operating parameter of the other cell system, in particular in the form of a control variable or controlled variable, are read out directly by means of the control unit of the other cell system, preferably provided by the control unit of the other cell system.Preferably, in a method step, in particular in a further data processing step, a data-based model for the cell system is trained based on data of at least one system behavior characteristic parameter of other cell systems, preferably based on data about at least one system behavior characteristic parameter of other cell systems and / or based on data about at least one system component characteristic parameter of at least one system component of other cell systems, in particular based on a data set of the data-based model. The training of the data-based model is based on, for example, linear regression, support vector machines, random forests, Gaussian processes, neural networks, interpretable boosting machines, etc. The trained data-based model is in particular configured to determine at least one calibration parameter based on at least one system component characteristic parameter of the cell system, preferably in a further data processing step. Advantageously, empirical values ​​of other cell systems can be used to parameterize the cell system. Advantageously, a particularly precise system parameterization can be achieved. Advantageously, a particularly efficient system regulation can be achieved.

[0014] It is also proposed that at least one calibration parameter is determined based on other models of the cell system (preferably other models already mentioned above, in particular different from the models) according to at least one system component characteristic parameter of the cell system and according to at least one system behavior characteristic parameter of the cell system. For example, the other model is constructed as a data-based model, and the model is constructed as a deterministic model, in particular a physical model. However, alternatively, it is also conceivable to construct the other model as a deterministic model, in particular a physical model, and to construct the model as a data-based model. In addition, it is alternatively conceivable to use only one model in the method, in particular a data-based model or a deterministic model, preferably a physical model. Advantageously, the determined calibration parameters can be checked particularly reliably.

[0015] It is also proposed that the result value of the calibration parameter is determined by comparing the calibration parameter value determined based on the model with the calibration parameter value determined based on other models. It is conceivable that the result value of the calibration parameter is determined based on the average value of the calibration parameter value determined based on the model and the calibration parameter value determined based on other models. In addition, it is also conceivable that the calibration parameter value determined based on the model and / or the calibration parameter value determined based on other models are provided with a weighting factor. However, alternatively, it is also conceivable that other calculation methods that are considered reasonable by a person skilled in the art are used to determine the result value of the calibration parameter based on the calibration parameter value determined based on the model and the calibration parameter value determined based on other models. In a method step, in particular in a parameterization step, at least one system component of the battery system, in particular a control unit, is calibrated using the result value of the calibration parameter, using the calibration parameter value determined based on the model or using the calibration parameter value determined based on other models. Advantageously, the calibration parameter can be determined particularly accurately. Advantageously, the battery system, in particular at least one system component, can be calibrated particularly accurately. Advantageously, a particularly effective system regulation of the battery system can be achieved.

[0016] It is further provided that the model is a deterministic model, in particular the deterministic model already mentioned above, and the other model is a data-based model, in particular the data-based model already mentioned above. Preferably, in a method step, in particular in a comparison step, the result value of the calibration parameter is determined by comparing the calibration parameter value determined on the basis of the deterministic model, preferably the physical model, with the calibration parameter value determined by means of the data-based model. Advantageously, the determined calibration parameter can be checked particularly reliably.

[0017] It is further proposed that aging characteristic parameters of the battery cell system are taken into account when determining the calibration parameters. It is conceivable that the input variables of the deterministic model (especially the physical model) include aging characteristic parameters of the battery cell system. The aging characteristic parameters especially describe the aging state of the battery cell system. For example, the aging characteristic parameter has a value of 1 for an unaged state of the battery cell system and preferably has a value greater than 1 for an aged state of the battery cell system. The aging characteristic parameter is preferably a scalar. The deterministic model, preferably the physical model, has clear reversibility in particular with respect to the aging characteristic parameters. It is conceivable that the deterministic model (especially the physical model) has clear reversibility with respect to the aging characteristic parameters, at least by filtering out physically irrelevant results of the deterministic model (especially the physical model). For example, a deterministic model (particularly a physical model) can be based on a model according to a publication by Zaccaria, Tucker and Traverso entitled "A distributed real-time model of degradation in a solid oxide fuel cell" (2016, Journal of Power Sources, 311, 175-181), which describes the degradation of a cell system with a solid oxide fuel cell, wherein the ohmic part of the internal voltage loss is provided with an aging characteristic parameter as a pre-factor. Alternatively, it is also conceivable, for example, that other resistances different from the ohmic part of the internal voltage loss of the cell system (particularly concentration losses or polarization losses) or the total internal resistance of the cell system are multiplied by the aging characteristic parameter as a pre-factor. Particularly preferably, in a deterministic model (particularly a physical model), the total specific surface resistance of the cell system is multiplied by the aging characteristic parameter. The deterministic model, in particular the physical model, preferably represents a mapping that maps input variables of the physical model (preferably at least one calibration parameter, at least one operating parameter of the cell system, at least one system component characteristic parameter of the cell system, and an aging characteristic parameter of the cell system) to output variables of the physical model (preferably at least one output characteristic parameter of the cell system). Preferably, the model, in particular the deterministic model, preferably the physical model, is inverted and optimized, preferably with respect to the aging characteristic parameter and at least one calibration parameter, to determine the aging characteristic parameter and at least one calibration parameter. Based on the deterministic model, in particular the physical model, preferably at least one calibration parameter and an aging characteristic parameter of at least one system behavior characteristic parameter of the cell system are determined from a data set of the deterministic model (in particular the physical model) and at least one system component characteristic parameter of the cell system, in particular in a data processing step.For example, based on a deterministic model, in particular a physical model, preferably in a data processing step, at least one calibration parameter and an aging characteristic parameter are determined, for which the deterministic model (in particular the physical model) together with at least one operating parameter of the cell system from the data set of the deterministic model (in particular the physical model) and at least one system component characteristic parameter of the cell system predicts at least one output characteristic parameter of the cell system from the data set of the deterministic model (in particular the physical model). In particular, in order to determine the aging characteristic parameter and at least one calibration parameter, preferably in the data processing step, an optimization method is applied, in particular the optimization methods already mentioned above, such as Bayesian optimization, simulated annealing methods, gradient methods, etc. It is also conceivable that the regression model is configured, in particular trained, to determine the aging characteristic parameter and at least one calibration parameter. Advantageously, the system regulation can be adapted to the aging state of the cell system. Advantageously, the cell system can be particularly effectively regulated according to the aging state of the cell system.

[0018] Furthermore, a cell system, in particular the cell system already mentioned above, is proposed, which has at least one electrochemical cell (in particular the electrochemical cell already mentioned above) and at least one control unit (in particular the control unit already mentioned above), which is used to perform the method according to the invention. A cell system with particularly precise control calibration can advantageously be provided. Advantageously, a particularly effective control regulation can be provided for the cell system. Advantageously, a particularly effective cell system can be provided.

[0019] Furthermore, a computer program is proposed, which comprises instructions which, when executed by a computer, cause the computer to perform a method according to the invention. The computer program is preferably stored on a storage element of a control unit. Alternatively, it is also conceivable to store the computer program on a portable data carrier, for example on a USB stick, on a portable hard disk, on an optical data carrier (in particular a CD etc.), in the cloud, on a server etc. In particular, the control unit instructs the computer to execute the computer program. Advantageously, the method for determining at least one calibration parameter of at least one system component of a battery cell system can be particularly conveniently and flexibly adapted. The method can advantageously be applied to different battery cell systems and in particular can be particularly conveniently adapted to the corresponding battery cell system.

[0020] The method according to the invention, the battery system according to the invention and / or the computer program according to the invention are not to be limited to the above-described applications and embodiments in this document. In particular, the method according to the invention, the battery system according to the invention and / or the computer program according to the invention may have a number of individual elements, components and units and method steps that differs from the number described herein in order to achieve the mode of action described herein. In addition, in the case of value ranges specified in the present disclosure, the values ​​within the limits mentioned should also be regarded as disclosed and can be used arbitrarily. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Further advantages emerge from the following description of the drawings. The drawings show embodiments of the invention. The drawings, the description and the claims contain numerous combinations of features. A person skilled in the art will also consider the features individually and combine them into further meaningful combinations.

[0022] Figure 1 A cell system according to the invention is shown, which has at least one electrochemical cell and at least one control unit for executing the method according to the invention for determining at least one calibration parameter of a system component of the cell system, and

[0023] Figure 2 A schematic flow of the method according to the present invention is shown. DETAILED DESCRIPTION

[0024] Figure 1 A cell system 22 is shown having a plurality of electrochemical cells 24 (in Figure 1 Only one electrochemical cell 24 is shown by way of example). Alternatively, it is also conceivable that the cell system 22 has only one electrochemical cell 24. The electrochemical cells 24 are combined into stacks. It is also conceivable that the cell system 22 includes a plurality of stacks formed by a plurality of electrochemical cells 24, the stacks preferably being combined into modules or towers. The electrochemical cells 24 are connected in series. The electrochemical cells 24 are configured as solid oxide fuel cells. Alternatively, it is also conceivable that the electrochemical cells 24 are configured as other fuel cells, for example, as polymer electrolyte fuel cells, etc., electrolyte cells or storage cells. For example, it is also conceivable that the electrochemical cells 24 are used in electrolyzers as an alternative.

[0025] The cell system 22 has two system components 30 , 32 . Alternatively, however, it is also conceivable that the cell system 22 has only one system component 30 , 32 or has more than two system components 30 , 32 .

[0026] The cell system 22 has at least one control unit 26. The system component 32 of the system components 30, 32 corresponds to the control unit 26. Other system components 30 of the system components 30, 32 are designed, for example, as fans, in particular air fans or circulation fans, mass flow controllers, heat exchangers, flow sensor systems, soft sensors or other system components of the cell system 22 that appear to be appropriate to a person skilled in the art. The control unit 26 is configured to execute a method for determining at least one calibration parameter of the system component 32 of the cell system 22, in particular the control unit 26.

[0027] Alternatively, however, it is also conceivable to provide that the control unit 26 is provided for determining at least one calibration parameter of a further system component 30 of the cell system 22, or for determining at least one calibration parameter of a further system component 30 of the cell system 22 and for determining at least one calibration parameter of a system component 32 of the cell system 22. For example, it is alternatively conceivable that the at least one calibration parameter is provided, for example, for calibrating a flow sensor system for air or fuel gas, air flow control, fuel gas flow control, temperature regulation, power regulation, fan control, soft sensors, etc.

[0028] The control unit 26 is configured to determine at least one calibration parameter of a system component 32 of the cell system 22 based on the model according to at least one system component characteristic parameter of the cell system 22 and according to at least one system behavior characteristic parameter of the cell system 22. The control unit 26 is configured to determine at least one calibration parameter of the system component 32 based on a further model according to at least one system component characteristic parameter of the cell system 22 and according to at least one system behavior characteristic parameter of the cell system 22. Alternatively, it is also conceivable to use only the model or the further model in the method.

[0029] The other system components 30 of the cell system 22 have at least one system component characteristic parameter of the cell system 22. The at least one system component characteristic parameter of the cell system can be, for example, a flow characteristic parameter of a fan (in particular a flow characteristic parameter of an air fan or a flow characteristic parameter of a circulation fan), a valve position of a valve (in particular a valve position of a needle valve), a pressure characteristic parameter of a fan (in particular a characteristic parameter related to a pressure buildup at the fan or a characteristic parameter related to a pressure reduction at the fan), a pressure characteristic parameter of a valve (in particular a characteristic parameter related to a pressure buildup at the valve or a characteristic parameter related to a pressure reduction at the valve), etc. Alternatively, however, it is also conceivable that the system component 32 of the cell system 22 has the system component characteristic parameter of the cell system 22, or that the system component 32 of the cell system 22 and the other system components 30 of the cell system 22 each have at least one system component characteristic parameter of the cell system 22, according to which the calibration parameters of the system component 32 of the cell system 22 are determined, in particular based on the model and / or other models together with at least one system behavior characteristic parameter of the cell system 22.

[0030] The control unit 26 has at least one processor (not shown here) and a storage element (not shown here) and an operating program stored on the storage element. The storage element is designed as a digital storage medium, such as a hard disk or the like.

[0031] The cell system 22 comprises a housing (not shown here) in which a stack formed by a plurality of electrochemical cells 24 is arranged. It is conceivable that the control unit 26 is at least partially arranged on the housing, in particular at least partially arranged in the housing. Alternatively, it is also conceivable that the cell system 22 has an external unit (not shown here), wherein the external unit comprises at least a part of the control unit 26. The external unit can be, for example, a cloud, a server, etc. It is also conceivable that the external unit comprises a storage element of the control unit 26. A computer program 28 is stored on the storage element, the computer program comprising instructions, which, when the computer program 28 is executed by a computer, causes the computer to perform a method for determining at least one calibration parameter of a system component 32, in particular of the control unit 26. The control unit 26 has a computer. Alternatively, it is also conceivable that the computer program 28 is stored on a portable data carrier, for example on a USB stick, on a portable hard disk, on an optical data carrier (in particular on a CD, etc.), in the cloud, on a server, etc.

[0032] The system behavior characteristic parameters include at least one output characteristic parameter of the cell system 22 and at least one operating parameter of the cell system 22. The output characteristic parameter of the cell system 22 is the output voltage of the cell system 22, in particular the output voltage of a stack formed by a plurality of electrochemical cells 24, electrical work, electrical power, etc. The at least one operating parameter of the cell system 22 is, for example, a requested load current, a temperature (in particular a gas temperature), a volume flow, an air utilization rate, a fuel utilization rate, a gas composition, a pressure, a pressure difference, etc. It is conceivable that the cell system 22 (in particular the control unit 26) or an external unit has a component detection unit, which is particularly used to detect (preferably in the form of a sensor) at least one component characteristic parameter of the cell system 22. Alternatively or additionally, it is conceivable that the cell system 22, preferably the system component 32 of the cell system 22 (in particular the control unit 26), or the external unit has an input unit for inputting at least one component characteristic parameter of the cell system 22. Preferably, the input unit is connected or connectable to the control unit 26 in terms of data technology, in particular for wireless and / or wired data transmission. The input unit may be configured as, for example, a keyboard, a keypad, a touch screen, a knob, a slide adjuster, a button, or the like.

[0033] The model is a deterministic model, in particular a physical model. The deterministic model, in particular the physical model is preferably a simulation model of the cell system 22, which preferably describes the operation of the cell system 22 in a steady state. Preferably, the deterministic model, in particular the physical model is based on a finite element method.

[0034] The deterministic model (preferably a physical model) has a plurality of input variables and at least one output variable. The input variables of the deterministic model (especially the physical model) include at least one system component characteristic parameter of the battery system 22. The input variables of the deterministic model (especially the physical model) have at least one calibration parameter. The input variables of the deterministic model (especially the physical model) include at least one operating parameter of the battery system 22.

[0035] When determining at least one calibration parameter, an aging characteristic parameter of the battery system 22 is taken into account. The input variables of the deterministic model, in particular the physical model, include the aging characteristic parameter of the battery system 22. The aging characteristic parameter describes the aging state of the battery system 22. For example, the aging characteristic parameter has a value of 1 for an unaged state of the battery system 22 and has a value greater than 1 for an aged state of the battery system 22. The aging characteristic parameter is a scalar. At least one output variable of the deterministic model, in particular the physical model, includes at least one output characteristic parameter of the battery system 22. Alternatively, it is also conceivable that the deterministic model, in particular the physical model, does not contain the aging characteristic parameter.

[0036] The deterministic model, preferably the physical model, has an unambiguous reversibility with respect to at least one calibration parameter and the aging characteristic parameter. It is conceivable that the deterministic model, preferably the physical model, has an unambiguous reversibility with respect to at least one calibration parameter and the aging characteristic parameter, at least by filtering out physically irrelevant results of the deterministic model, preferably the physical model.

[0037] For example, a deterministic model (particularly a physical model) can be based on a model according to a publication by Zaccaria, Tucker and Traverso entitled "A distributed real-time model of degradation in a solid oxide fuel cell" (2016, Journal of Power Sources, 311, 175-181), which describes the degradation of a cell system with a solid oxide fuel cell, wherein the ohmic part of the internal voltage loss is provided with an aging characteristic parameter as a pre-factor. Alternatively, it is also conceivable, for example, that other resistances of the internal voltage loss of the cell system 22 that are different from the ohmic part (particularly concentration losses or polarization losses) or the total internal resistance of the cell system 22 are multiplied by the aging characteristic parameter as a pre-factor. Particularly preferably, in a deterministic model (particularly a physical model), the total specific surface resistance of the cell system 22 is multiplied by the aging characteristic parameter.

[0038] A deterministic model (especially a physical model) represents a mapping of the input variables of the physical model (especially at least one calibration parameter, at least one operating parameter of the battery system 22, at least one system component characteristic parameter of the battery system 22 and an aging characteristic parameter of the battery system 22) to the output variables of the physical model, preferably a mapping of at least one output characteristic parameter of the battery system 22.

[0039] Figure 2 A schematic flow chart of a method for determining at least one calibration parameter of a system component 32 (in particular a control unit 26) of a cell system 22 is shown. Based on a model of the cell system 22, at least one calibration parameter is determined according to at least one system component characteristic parameter of the cell system 22 and according to at least one system behavior characteristic parameter of the cell system 22. Based on other models of the cell system 22, at least one calibration parameter is determined according to at least one system component characteristic parameter of the cell system 22 and according to at least one system behavior characteristic parameter of the cell system 22.

[0040] In one method step, in particular in the data generation step 10, at least one data set is generated for a deterministic model (in particular a physical model), which data set has data about at least one system behavior characteristic parameter of the cell system 22, preferably data about at least one output characteristic parameter of the cell system 22 and data about at least one operating parameter of the cell system 22. The data set of the deterministic model (in particular a physical model) is generated by means of a statistical experimental design, for example by means of Latin hypercube sampling, active learning, etc. The data set of the deterministic model (in particular a physical model) describes the target behavior of the cell system 22.

[0041] The data about at least one system behavior characteristic parameter of the battery system 22 from the data set of the deterministic model, in particular the physical model, are generated in such a way that the associated system state of the battery system 22 can be described by the deterministic model, in particular the physical model. For example, it is conceivable that the deterministic model, in particular the physical model, can be applied only to stable conditions, so that sufficiently stable system behavior characteristic parameters should be used accordingly. It is also conceivable to at least partially pre-process at least one system behavior characteristic parameter of the battery system 22, preferably at least one output characteristic parameter of the battery system 22 and / or at least one operating parameter of the battery system 22, for example, by removing oscillations through a filter, etc.

[0042] It is conceivable that at least one system behavior characteristic parameter of the cell system 22 is at least partially converted in a method step, in particular in a data processing step 12, to be adapted to a deterministic model, in particular a physical model, preferably by the control unit 26. For example, if necessary, at least one system behavior characteristic parameter of the cell system 22 from a data set of a deterministic model, preferably a physical model, is converted in the data processing step 12 into an input variable and / or output variable of the deterministic model, in particular a physical model.

[0043] Based on a deterministic model, in particular a physical model, at least one calibration parameter and an aging characteristic parameter are determined in a data processing step 12 from a data set of the deterministic model (in particular the physical model) and at least one system component characteristic parameter of the cell system 22 for at least one system behavior characteristic parameter of the cell system 22. Based on a deterministic model (in particular the physical model), at least one calibration parameter and an aging characteristic parameter are determined in a data processing step 12, for which at least one calibration parameter and aging characteristic parameter the deterministic model (in particular the physical model) together with at least one operating parameter of the cell system 22 from the data set of the deterministic model (in particular the physical model) and at least one system component characteristic parameter of the cell system 22 predicts at least one output characteristic parameter of the cell system 22 from the data set of the deterministic model (in particular the physical model). In order to determine the calibration parameter and in particular the aging characteristic parameter, an optimization method is applied in the processing step, in particular the optimization method already mentioned above, such as Bayesian optimization, simulated annealing method, gradient method, etc. The deterministic model, in particular the physical model, inverts and optimizes the model (preferably the deterministic model, preferably the physical model). In order to determine the calibration parameters and in particular the aging characteristic parameters, a model (preferably a deterministic model, preferably a physical model) is inverted and optimized for at least one calibration parameter and preferably for the aging characteristic parameters. In order to determine at least one calibration parameter and in particular the aging characteristic parameter in the data processing step 12, an optimization method is applied, preferably to the inverted model, such as Bayesian optimization, simulated annealing methods, gradient methods, etc.

[0044] It is also conceivable to train the regression model using a deterministic model, in particular a physical model. Preferably, the regression model is trained based on the results determined with the aid of the physical model. The trained regression model is configured to replace the deterministic model, in particular the physical model, to at least be used to determine at least one calibration parameter and an aging characteristic parameter based on at least one system behavior characteristic parameter of the battery system 22 and based on at least one system component characteristic parameter of the battery system 22, in particular in the data processing step 12. The regression model is configured to be used in the optimization method to determine at least one calibration parameter and an aging characteristic parameter or to directly determine at least one calibration parameter and an aging characteristic parameter based on at least one system behavior characteristic parameter of the battery system 22 and at least one system component characteristic parameter of the battery system 22. The training of the regression model is based on, for example, linear regression, random forest, Gaussian process, neural network, interpretable boosting machine, etc.

[0045] The other model is a data-based model. As data of a system behavior characteristic parameter of the cell system 22, data of at least one system behavior characteristic parameter of at least one other cell system (not shown here) are used at least partially, the other cell system having at least one electrochemical cell, in particular when determining at least one calibration parameter and preferably an aging characteristic parameter based on the other model. At least one system behavior characteristic parameter of the other cell system is a variable similar to at least one system behavior characteristic parameter of the cell system 22. In a method step, in particular in the data provision step 14, a data set of the data-based model is generated. The data set of the data-based model includes data on at least one system behavior characteristic parameter of the other cell system, preferably data on at least one output characteristic parameter of the other cell system and data on at least one operating parameter of the other cell system, which data are detected on the other cell system during operation of the other cell system. The data set of the data-based model includes data on at least one system component characteristic parameter of a system component of the other cell system. At least one operating parameter of the other cell system is a variable similar to at least one operating parameter of the cell system 22. At least one output characteristic parameter of the other cell system is a variable similar to at least one output characteristic parameter of the cell system 22.

[0046] It is conceivable that the other battery system has a detection unit for at least partially detecting data about at least one system behavior characteristic parameter of the other battery system at the other battery system. The detection unit is preferably connected to the control unit 26 of the battery system 22 in terms of data technology, in particular to an external unit of the battery system 22, preferably for wireless and / or wired data transmission. It is conceivable that data about at least one output characteristic parameter of the other battery system and / or data about at least one operating parameter of the other battery system are detected at the other battery system, preferably in a sensor manner, preferably with the aid of the detection unit. The detection unit comprises, for example, at least one sensor unit for at least partially detecting data about at least one system behavior characteristic parameter of the other battery system at the other battery system. The detection unit is preferably configured to detect at least data about the output characteristic parameter of the other battery system at the other battery system, in particular in a sensor manner. The sensor unit comprises, for example, at least one voltage sensor, in particular for detecting data about the output characteristic parameter of the other battery system, which is configured as an output voltage, at the other battery system. It is conceivable that the detection unit is configured to at least partially detect data about at least one operating parameter of the other battery system at the other battery system. Additionally or alternatively, it is conceivable that data about at least one operating parameter of the other cell system, in particular in the form of a control variable or controlled variable, are read out directly by means of the control unit of the other cell system, preferably provided by the control unit of the other cell system.

[0047] In one method step, in particular in a further data processing step 16, a data-based model for the cell system 22 is trained based on data of at least one system behavior characteristic parameter of other cell systems, preferably based on data about at least one system behavior characteristic parameter of other cell systems and based on data about at least one system component characteristic parameter of at least one system component of other cell systems, in particular based on a data set of a data-based model. The training of the data-based model is based, for example, on linear regression, support vector machines, random forests, Gaussian processes, neural networks, interpretable boosting machines, etc. In a further data processing step 16, at least one calibration parameter is determined based on at least one system component characteristic parameter of the cell system 22 with the aid of the trained data-based model.

[0048] In one method step, in particular in the evaluation step 18, a result value of the calibration parameter is determined by comparing a calibration parameter value determined based on a model, in particular a deterministic model, preferably a physical model, with a calibration parameter value determined based on another model, in particular a data-based model, preferably with the aid of a control unit 26.

[0049] It is conceivable that the result value of the calibration parameter is determined based on the average value of the calibration parameter determined based on the model and the calibration parameter value determined based on other models. In addition, it is also conceivable that the calibration parameter value determined based on the model and / or the calibration parameter value determined based on other models are provided with a weighting factor. However, alternatively, it is also conceivable that other calculation methods that are considered reasonable by those skilled in the art are used to determine the result value of the calibration parameter based on the calibration parameter value determined based on the model and the calibration parameter value determined based on other models.

[0050] In one method step, in particular in the parameterization step 20 , at least one system component 32 , in particular the control unit 26 , is calibrated using the determined calibration parameters.

Claims

1. A method for determining at least one calibration parameter of at least one system component (30, 32), in particular a control unit (26), of a cell system (22), the cell system having at least one electrochemical cell (24), characterized in that: The calibration parameter is determined based on at least one model of the battery system (22), according to at least one system component characteristic parameter of the battery system (22) and according to at least one system behavior characteristic parameter of the battery system (22).

2. The method according to claim 1, characterized in that The model is inverted and optimized in order to determine the calibration parameters.

3. The method according to claim 1 or 2, characterized in that: The model is a deterministic model used to train a regression model.

4. The method according to any one of the preceding claims, characterized in that As data of at least one system behavior characteristic parameter of the cell system (22), data of at least one system behavior characteristic parameter of at least one other cell system having at least one electrochemical cell are at least partially used.

5. The method according to any one of the preceding claims, characterized in that Based on other models of the battery system (22), in particular other models different from the model, the calibration parameter is determined according to at least one system component characteristic parameter of the battery system (22) and according to at least one system behavior characteristic parameter of the battery system (22).

6. The method according to claim 5, characterized in that A resulting value of the calibration parameter is determined by comparing the calibration parameter value determined based on the model with the calibration parameter value determined based on the other model.

7. The method according to claim 5 or 6, characterized in that: The model is a deterministic model and the other model is a data-based model.

8. The method according to any one of the preceding claims, characterized in that When determining the calibration parameters, an aging characteristic parameter of the cell system (22) is taken into account.

9. A cell system (22) having at least one electrochemical cell (24) and at least one control unit (26) for carrying out the method according to one of the preceding claims.

10. A computer program (28) comprising instructions which, when the computer program (28) is executed by a computer, cause the computer to perform the method according to one of claims 1 to 8.

Images