Method for correcting fuel concentration estimates, fuel cell system, vehicle, computer program product and storage medium

Abstract

The present invention relates to a method for correcting fuel concentration estimates in an anode (13) of a fuel cell system (10), comprising: determining fuel concentration sensor values ​​in an exhaust gas path (18) using a fuel sensor (20) positioned in the exhaust gas path (18), calculating fuel concentration model values ​​in the exhaust gas path (18) using fuel concentration estimates, performing comparisons between the fuel concentration sensor values ​​and the fuel concentration model values, and correcting the fuel concentration estimates based on the comparisons. The invention further relates to a fuel cell system (10), a vehicle (100), a computer program product (30), and a computer-readable storage medium (31).

Description

[0001] The present invention relates to devices and methods for correcting fuel concentration estimates in a fuel cell system. In particular, the invention relates to a method for correcting fuel concentration estimates in a fuel cell system, a fuel cell system configured and designed to carry out the method, and a vehicle having such a fuel cell system. The invention further relates to a computer program product for carrying out the method and a computer-readable storage medium on which such a computer program product is stored.

[0002] Known fuel cell systems typically comprise a stacked arrangement with a cathode and an anode, a cathode inlet path for conducting oxidant into the cathode, a cathode outlet path for conducting cathode exhaust gas from the cathode, an anode inlet path for conducting fuel into the anode, and an anode outlet path for conducting anode exhaust gas from the anode. Furthermore, fuel cell systems are known to have an exhaust path through which the cathode exhaust gas and the anode exhaust gas can be discharged together into the environment of the fuel cell system. For the most efficient and safe operation of the fuel cell system, it is important to always be aware of the fuel concentration in the anode. Various options are known for determining the fuel concentration.For example, systems are known in which the fuel concentration is determined using a sensor positioned in the anode. Furthermore, systems are known in which the fuel concentration is estimated or approximately calculated based on the shutdown time of the fuel cell system in combination with a system model. In the latter case, it is known to calculate the desired fuel cell concentration using a virtual sensor in the anode and, in particular, to correct an estimated fuel concentration value. The known methods and devices for correcting the estimated values ​​are relatively complex and yet often do not lead to the desired result.

[0003] The object of the present invention is to provide improved methods and devices for calculating fuel concentration values ​​and / or for correcting estimated fuel concentration values ​​in the anode of a fuel cell system.

[0004] The above object is achieved by the patent claims. In particular, the above object is achieved by the method according to claim 1 and by the fuel cell system, the vehicle, the computer program product, and the computer-readable storage medium according to the independent claims. Further advantages of the invention emerge from the dependent claims, the description, and the figures. Features described in connection with the method also apply in connection with the method according to the invention, the fuel cell system according to the invention, the vehicle according to the invention, the computer program product according to the invention, the storage medium according to the invention, and vice versa, so that with regard to the disclosure of the individual aspects of the invention, reference is and / or can always be made to each other.

[0005] According to a first aspect of the present invention, a method is proposed for correcting fuel concentration estimates of a fuel concentration in an anode of a fuel cell system, wherein the fuel cell system comprises a stack assembly with a cathode and the anode, a cathode inlet path for conducting oxidant into the cathode, a cathode outlet path for conducting cathode exhaust gas from the cathode, an anode inlet path for conducting fuel into the anode, an anode outlet path for conducting anode exhaust gas from the anode, an exhaust path for conducting the cathode exhaust gas and the anode exhaust gas into the environment of the stack assembly, and an anode purge valve for selectively conducting anode exhaust gas from the anode outlet path toward the exhaust path. The method comprises the following steps: - Determining fuel concentration sensor values ​​in the exhaust gas path using a fuel sensor positioned in the exhaust gas path, - Calculating fuel concentration model values ​​in the exhaust gas path using fuel concentration estimates, - performing comparisons between the fuel concentration sensor values ​​and the fuel concentration model values, and - Correct the fuel concentration estimates based on the comparisons.

[0006] The proposed method makes it possible to correct existing fuel concentration estimates in a simple yet relatively precise manner. This allows for the prevention of undersupply and thus potential damage to the fuel cell system, even without a specific fuel sensor in the anode or anode subsystem. Consequently, operation of the anode with an unnecessarily high fuel concentration can also be avoided. In other words, an improved operating point can be achieved, preventing fuel losses and the resulting lower fuel and system efficiency. The fuel concentration estimates in the anode can also be understood as fuel concentration estimates in the anode subsystem upstream of the exhaust path.Accordingly, the method can in principle also be used to correct fuel concentration estimates of a fuel concentration in the anode subsystem upstream of the exhaust gas path.

[0007] The aforementioned advantages are achieved in particular by the fuel sensor positioned in the exhaust gas path and the proposed method steps according to which the measurement data from the fuel sensor is processed. In conventional systems and methods, the physical and / or virtual sensors were positioned in the area of ​​the anode. Alternatively, this was omitted entirely. Within the scope of the invention, it was recognized that by positioning a physical sensor in the exhaust gas path, i.e. downstream of the anode outlet path and / or an anode subsystem, sufficiently precise conclusions can still be drawn about the fuel concentration in the anode outlet path and / or in the anode itself. Furthermore, it has been shown that positioning the fuel sensor in the exhaust gas path can bring further advantages with regard to improved installation space design of the fuel cell system.

[0008] Furthermore, it is advantageous that the proposed method can reliably meet legal requirements for determining potential fuel emissions into the environment of the fuel cell system, for example, into the environment of a vehicle with a fuel cell system. Based on the comparisons and the sensor data, relatively accurate and reliable statements about the emitted fuel can be made.

[0009] Furthermore, the application of this method reduces the requirements for sensor technology. For example, relatively few measurement methods are permitted in the anode, since conventional fuel cell stacks and the entire anode are considered to be an EX zone, meaning a zone in which ignition sources should be kept to a minimum.

[0010] The proposed process steps can be performed continuously during operation of the fuel cell system. If a relatively large difference still exists between the fuel concentration sensor values ​​and the fuel concentration model values ​​at the beginning of the process, the initially used fuel concentration estimate will become closer to the actual fuel concentration value in the anode with each comparison and consequently with each correction. This means that even with a relatively high initial value error, the proposed process makes it possible to relatively quickly bring the fuel concentration estimate close to the actual fuel concentration value in the anode, or, based on this, to operate the fuel cell as efficiently and as gently as possible.

[0011] The fuel concentration estimate can be understood as a value for a fuel concentration within the anode and / or in an anode subsystem. In this case, the fuel can be understood to mean, in particular, hydrogen. The fuel can also be understood to mean a hydrocarbon or a hydrocarbon mixture, for example, natural gas. In addition to the fuel, other process gases such as nitrogen or water vapor can be present in and / or at the anode.

[0012] Determining the fuel concentration sensor values ​​in the exhaust gas path can be understood to mean that the fuel concentration in the exhaust gas path is determined using the fuel sensor. The same applies analogously to the further method steps. Determining the fuel concentration sensor values ​​using the fuel sensor can be understood to mean that the fuel concentration sensor values ​​are determined and / or measured directly, indirectly and / or exclusively by the fuel sensor. It is preferred that measured values ​​are obtained by the fuel sensor, based on which the fuel concentration or the corresponding fuel concentration sensor values ​​are determined accordingly indirectly. Determining the fuel concentration sensor values ​​can therefore be understood to mean that the fuel concentration sensor values ​​are determined based on measured values ​​and calculations using the measured values.

[0013] The fuel sensor used can, for example, be configured to determine the heat transfer of the passing exhaust gas and / or generate corresponding measured values. In dry environments, this can be used to determine the fuel concentration relatively precisely. In the event that such a measuring principle reacts sluggishly to the dynamics and water vapor content from the cathode, an offset correction value can be used, which can be subtracted and / or included for the fuel sensor depending on the fuel cell system in question.

[0014] Correcting the fuel concentration estimates can be understood as determining and / or calculating at least one correction value for an initially assumed fuel concentration estimate, wherein the correction value can be used as a new or future fuel concentration estimate for the further calculation of another fuel concentration model value. This means that the proposed method is preferably carried out by calculating a first fuel concentration model value based on an initially used fuel concentration estimate, based on which the initially used fuel concentration estimate is corrected, in order to subsequently use this corrected fuel concentration estimate to calculate another fuel concentration model value, based on which the first corrected fuel concentration estimate is corrected again.This process can be continued throughout the entire operation of the fuel cell system, especially cyclically. Correction may not always be necessary during a single operation of the fuel cell system. If a fuel concentration sensor value matches a calculated fuel concentration model value, the correction of the estimated fuel concentration can be omitted, at least during this cycle.

[0015] The stack arrangement can comprise at least one fuel cell or multiple fuel cells in the form of a fuel cell stack. The anode purge valve can be configured to purge a so-called anode subsystem. An anode purge valve can also be referred to as a purge valve. By repeatedly opening and closing the anode purge valve, anode exhaust gases can be extracted from a recirculation path of the anode subsystem and introduced into the exhaust path and / or an anode purge line. The anode purge valve can be positioned in, on, or downstream of a water separator of the anode subsystem. The anode purge valve can be designed as an electromagnetic valve.

[0016] The fuel concentration model values ​​are preferably performed in a temporally correlated manner with a purging process of the fuel cell system, i.e., with the activation and deactivation of the anode purge valve. The fuel concentration model values ​​can therefore be calculated cyclically when the anode purge valve is activated, and the fuel concentration estimates corrected based on them can be corrected cyclically. During a cycle, the fuel concentration model value can be calculated via multiple conversions based on various physical variables, the geometry of the stack arrangement, and / or the operating state of the fuel cell system. During the purge process, reaction off-gases such as liquid water, nitrogen, water vapor, and gaseous fuel, in particular hydrogen, are discharged from the anode.This purge discharge corresponds to the molar ratios at the anode outlet and can therefore be used to draw conclusions about the current fuel concentration in the anode.

[0017] The fuel cell system may comprise an anode subsystem formed by the fuel-carrying components of the fuel cell system. The main function of the anode subsystem is to supply and distribute fuel to the electrochemically active surfaces of the anode, to recirculate and / or recycle unused fuel, and to remove, particularly cyclically remove, anode exhaust gas into the exhaust path.

[0018] Determining the fuel concentration sensor values, calculating the fuel concentration model values, performing the comparisons and / or correcting the fuel concentration estimates can be performed online on at least one control unit and / or computer of the fuel cell system.

[0019] According to a further embodiment of the present technology, it is possible that: - the fuel concentration sensor values ​​are determined at least partially during discharge of anode exhaust gas into the exhaust path and / or that the fuel concentration model values ​​are calculated at least partially during discharge of anode exhaust gas into the exhaust path, - when determining the fuel concentration sensor values, lower sensor values ​​for a closed state of the anode purge valve, upper sensor values ​​for an open state of the anode purge valve and sensor difference values ​​are determined between a lower sensor value and an upper sensor value, - when calculating the fuel concentration model values, lower model values ​​for a closed state of the anode purge valve, upper model values ​​for an open state of the anode purge valve and model difference values ​​are determined between a lower sensor value and an upper sensor value, - when performing the comparisons, the sensor difference values ​​are compared with the model difference values.

[0020] By comparing the various difference values, the fuel concentration estimates can be corrected relatively easily and reliably toward the desired real value. When calculating the respective difference values, a difference can be calculated between a minimum concentration value determined in the respective cycle and a maximum concentration value determined in the same cycle. In other words, a difference value can be calculated between a valley and a peak of a measurement graph. When performing the comparisons, or through the comparisons, a sensor / model difference value can then be calculated between a model difference value of a cycle and a sensor difference value of the same cycle.This sensor / model difference value can then be used to correct the currently available fuel concentration estimate, i.e. to include it in the calculation.

[0021] The proposed procedure also allows: - a fuel concentration estimate is corrected upwards if a comparison between a sensor difference value and a model difference value shows that the sensor difference value is greater than the model difference value and / or - a fuel concentration estimate is corrected downwards if a comparison between a sensor difference value and a model difference value shows that the sensor difference value is smaller than the model difference value.

[0022] Accordingly, it is proposed that instead of attempting to correct the current fuel concentration estimate as directly as possible to the actual fuel concentration value in the anode, an iterative approximation should be used. This approach allows for, in particular, measurement inaccuracies and / or signal noise during operation of the fuel cell system to be taken into account. A fuel concentration value that corresponds to the actual fuel concentration value, or at least is relatively close to it, can still be determined quickly and reliably.

[0023] Furthermore, in a method according to the present technology, it is possible for the estimated fuel concentration to be corrected upwards and / or downwards by less than 10%. As already mentioned above, the current estimated fuel concentration should not be corrected as directly as possible to the actual fuel concentration value within the framework of the method, but rather via an iterative approximation to the actual fuel concentration value. By limiting the rate of change to a value of less than 10% or less than 5%, measurement inaccuracies and / or signal noise during operation of the fuel cell system can be taken into account particularly reliably. It can also be advantageous if the estimated fuel concentration is corrected upwards and / or downwards by less than 4%, less than 3%, less than 2%, or less than 1% when carrying out the method.

[0024] Furthermore, in a method according to the present technology, it is possible for the fuel concentration model values ​​to be calculated in a temporally correlated manner to the fuel concentration sensor values ​​and, in doing so, each with at least one predefined time delay value. In this way, the fuel concentration model values ​​can be calculated in such a way that a comparison with the fuel concentration sensor values ​​that is as meaningful and / or simple as possible can subsequently be carried out. The at least one predefined time delay value can have a time value in a range between 400 ms and 800 ms, for example approximately 600 ms. Such a delay value can take into account the distance in the exhaust path that the anode exhaust gas and / or the exhaust gas must travel before reaching the fuel sensor.Depending on the position of the fuel sensor, the geometry of the exhaust path, and / or the presence of other components in the exhaust path, such as a silencer, the delay value can be selected differently. The at least one predefined time delay value can additionally have a time value in a range between 30 ms and 80 ms, for example, approximately 50 ms. Such a delay value can take into account the distance in a purge path that the anode exhaust gas must travel from the anode purge valve until it reaches the exhaust path. Depending on the position of the anode purge valve and / or the geometry of the anode subsystem, this delay value can also be selected and / or used differently.

[0025] In a method according to the present technology, it is also possible to use a respective actuation point of the anode purge valve, at which the anode purge valve is actuated to direct anode exhaust gas from the anode outlet path toward the exhaust path, to calculate the respective fuel concentration model value. As already mentioned above, it may be advantageous to calculate the respective fuel concentration model values ​​whenever the anode purge valve is actuated and / or after the anode purge valve has been actuated. By taking the actuation point(s) into account, fuel concentration model values ​​can be calculated particularly accurately.

[0026] Another aspect of the proposed technology relates to a fuel cell system comprising a stack assembly with a cathode and an anode, a cathode inlet path for conducting oxidant into the cathode, a cathode outlet path for conducting cathode exhaust gas from the cathode, an anode inlet path for conducting fuel into the anode, an anode outlet path for conducting anode exhaust gas from the anode, an exhaust gas path for conducting the cathode exhaust gas and the anode exhaust gas into the environment of the stack assembly, an anode purge valve for selectively conducting anode exhaust gas from the anode outlet path toward the exhaust gas path, and a fuel sensor positioned in the exhaust gas path, wherein the fuel cell system is configured and designed to carry out a method as described above.

[0027] The fuel cell system according to the invention thus offers the same advantages as those described in detail with reference to the method according to the invention. The fuel cell system is preferably configured for mobile applications such as vehicles, in particular for providing energy for at least one drive unit, such as an electric motor, for propelling the vehicle. The stack arrangement can be designed in the form of a fuel cell stack with one or more fuel cells. In its simplest form, the fuel cell is an electrochemical energy converter that converts fuel and oxidant into reaction products, thereby producing electricity and heat. The anode and cathode can be separated from one another by an ion-selective or ion-permeable separator.To carry out the method, the fuel cell system may have a suitable determination and / or calculation unit, for example, as part of at least one computer and / or at least one control unit. The computer and / or the control unit may further have comparison units and / or correction units for performing the comparisons and / or corrections.

[0028] A further aspect of the invention relates to a vehicle having a fuel cell system as described above and at least one electric motor for driving the vehicle, wherein the fuel cell system is configured to supply power to the at least one electric motor. The vehicle thus offers the same advantages as have been described in detail with reference to the fuel cell system. The vehicle can be understood to mean a motor vehicle such as a motor-driven two-wheeler, a car, or a truck. The vehicle can also be understood to mean a road vehicle, an aircraft, a watercraft, a rail vehicle, a spacecraft, or a robot. The vehicle can also be understood to mean a purely electric vehicle or a hybrid electric vehicle, which, in addition to the at least one electric motor, has an internal combustion engine for driving the vehicle.

[0029] Furthermore, the invention comprises a computer program product and a computer-readable, in particular non-volatile, storage medium on which the computer program product is stored. Thus, the computer program product and the computer-readable storage medium also provide the advantages described above. The computer program product can include instructions which, when the computer program product is executed by a computer, for example, a vehicle control unit, cause the computer to execute the proposed method in a vehicle as described above. In a broader sense, the computer-readable storage medium can also be understood to include a control unit, for example, a vehicle control unit, with the computer program product installed therein.

[0030] The computer program product may be implemented as computer-readable instruction code in any suitable programming language and / or machine language, such as JAVA, C++, C#, and / or Python. The computer program product may be stored on a computer-readable storage medium, such as a data disk, a removable drive, volatile or non-volatile memory, or a built-in memory / processor. The instruction code may program a computer or other programmable device, such as a control unit, to perform the desired functions. Furthermore, the computer program product may be provided and / or be provided on a network, such as the Internet, from which it can be downloaded by a user as needed.The computer program product can be and / or be implemented by means of software as well as by means of one or more special electronic circuits, i.e. in hardware or in any hybrid form, i.e. by means of software components and hardware components.

[0031] Further measures improving the invention will become apparent from the following description of various exemplary embodiments of the invention, which are schematically illustrated in the figures. All features and / or advantages apparent from the claims, the description, or the figures, including structural details and spatial arrangements, may be essential to the invention both individually and in various combinations.

[0032] They show schematically: Fig. 1 a fuel cell system according to an embodiment of the present technology, Fig. 2 a vehicle according to an embodiment of the present technology, Fig. 3 a computer-readable storage medium having a computer program product stored thereon according to an embodiment of the present technology, Fig. 4 is a flowchart for explaining a method according to a first embodiment of the present technology, Fig. 5 is a flowchart for explaining a method according to a second embodiment of the present technology, and Fig. 6 a characteristic diagram to further explain the procedures, and Fig. 7 shows another characteristic diagram to further explain the procedures.

[0033] Elements with the same function and mode of operation are provided with the same reference symbols in the figures.

[0034] Fig. 1 shows a fuel cell system 10 according to a possible embodiment. The illustrated fuel cell system 10 comprises a stack assembly 11 in the form of a fuel cell stack with a cathode 12 and an anode 13, a cathode inlet path 14 for conducting oxidant into the cathode 12, a cathode outlet path 15 for conducting cathode exhaust gas from the cathode 12, an anode inlet path 16 for conducting fuel into the anode 13, and an anode outlet path 17 for conducting anode exhaust gas from the anode 13. The fuel cell system 10 further comprises an exhaust path 18 for conducting the cathode exhaust gas and the anode exhaust gas into the environment of the stack assembly 11. For mixing the cathode exhaust gas with the anode exhaust gas, a mixing region 21 is designed, in which the anode outlet path 17 opens into the cathode outlet path 15.At this point it should be mentioned that the fuel cell system 10 and in particular the cathode subsystem of the fuel cell system 10 are shown in a highly simplified manner for a better overview.

[0035] The fuel cell system 10 shown has an anode purge valve 19 in the anode subsystem for selectively directing anode exhaust gas from the anode outlet path 17 toward the exhaust path 18. The fuel cell system 10 further has a blower 24 and a water separator 25 in the anode subsystem. As an alternative to the blower 24, it would also be possible to implement passive recirculation via an injector / ejector component. Furthermore, the fuel cell system 10 has an inlet valve 26 and a diffuser 27 in the anode subsystem. Furthermore, the fuel cell system 10 has numerous sensors and actuators. Shown are a fuel sensor 20 positioned in the exhaust path 18 for determining fuel concentration sensor values ​​in the exhaust path 18 and an air mass sensor 22 positioned in the cathode inlet path 14 for determining an air mass or air flow values ​​in the cathode inlet path 14.In addition, a pressure sensor 23 is positioned in the anode outlet path 17, more precisely upstream of the anode purge valve 19 and downstream of the anode 13. Using the pressure sensor 23, the gas pressure upstream of the anode purge valve 19 and thus approximately the pressure in the anode 13 and / or at the anode outlet of the stack assembly 11 can be determined.

[0036] The fuel sensor 20, the air mass sensor 22, and the pressure sensor 23 are signal-connected to a control unit 32. The control unit 32 is configured to control various actuators and valves, for example, the anode purge valve 19, based on the received sensor data. Furthermore, calculations can be performed in the control unit 32 based on the sensor data and, for example, on the basis of detected valve switching times of the anode purge valve 19. The control unit 32 can also be signal-connected to other functional components of the fuel cell system 10 shown, although a complete wiring and / or connection diagram has been omitted for the sake of clarity. Fig. The fuel cell system 10 shown in Figure 1 is configured, in particular by means of the sensors shown and the control unit 32, to correct fuel concentration estimates of a fuel concentration in the anode 13 of the fuel cell system 10. Furthermore, it is possible for the control unit 32, the sensors shown, and other functional components to transmit and / or receive signals wirelessly, rather than wired.

[0037] In Fig. 2 shows a vehicle 100 with a fuel cell system 10 as described above. The vehicle 100 also has two electric motors 40, which are positioned in the area of ​​the front axle and in the area of ​​the rear axle of the vehicle 100 as drive motors of the vehicle 100. In the example shown, the fuel cell system 10 is configured to supply power to the two electric motors 40 and is electrically connected to the electric motors 40 accordingly. The vehicle 100 further has a fuel tank 50 for supplying fuel to the fuel cell system 10.

[0038] Fig. Figure 3 shows a computer-readable and non-volatile storage medium 31 with a computer program product 30 stored thereon. In the example shown, the storage medium 32 is configured in the form of a memory stick or flash drive. The computer program product 30 comprises instructions that are executed when the computer program product 30 is executed by a computer, for example in the form of the Fig. 1, cause the control unit 32 to operate in the manner shown in Fig. 2 to perform a method for correcting fuel concentration estimates.

[0039] With reference to Fig. 4 and reference to Fig. 1, essential method steps of a method for correcting fuel concentration estimates of a fuel concentration in the anode 13 of the illustrated fuel cell system 10 are explained below. In a first step S1, a fuel concentration sensor value in the exhaust gas path 18 is determined using the fuel sensor 20. In a second step S2, a fuel concentration model value in the exhaust gas path 18 is calculated using a fuel concentration estimate. Subsequently, in a third step S3, a comparison is carried out between the fuel concentration sensor value and the fuel concentration model value. Subsequently, in a fourth step S4, the initial fuel concentration estimate is corrected based on the comparison. The method is then carried out again in order to determine the fuel concentration based on a further comparison orto further correct the fuel concentration estimate based on further comparisons or to keep it at a value as close as possible to the actual fuel concentration value.

[0040] With reference to the Fig. 5 and Fig. 6, a possible embodiment of the method for correcting fuel concentration estimates or for cyclically correcting a fuel concentration estimate is then described in detail.

[0041] In a first step S10, a fuel concentration sensor value in the exhaust gas path 18 is first determined using the fuel sensor 20. The fuel concentration sensor value can be determined as con(H2Real_ExhaustSensor) as a physical value, where "con" stands for the concentration of the fuel in the exhaust gas.

[0042] In a second step S20, a fuel concentration model value is calculated in the exhaust path 18 as con(H2Est_ExhaustSensor), where "Est" stands for estimation. The fuel concentration model value con(H2Est_ExhaustSensor) is calculated as a model value delayed by 600 ms in the mixing area 21 as con(H2Est_Mixing area, t-600ms). The delay value results from the length in the exhaust path 18 from the mixing area 21 to the fuel sensor 20.

[0043] In a third step S30, the model value con(H2Est_Mixing_Area) is calculated taking into account the volume flow of the exhaust air in the cathode outlet path 15 according to con(H2Est_Mixing_Area) = volf(H2Est_Purge_Mixing_Area) / volf(Stack_Exhaust_Air), where volf stands for Volume Flow. The fuel volume concentration can be determined or calculated from the ratio of the volume flows. For gases, it can be assumed that both the volume flow and the molar flow or flow can be used to determine the concentration. For ideal gas mixtures, the values ​​of volume fraction and volume concentration are the same. For ideal gas mixtures, there is also equality with the molar fraction and the particle number fraction.

[0044] In a fourth step S40, the volume flow of the exhaust air is calculated as volf(stack_exhaust_air) = volf(air_sensor) - volf(stack_consumption) - volf(air_losses). This means that the volume flow of the stack exhaust air is determined using a physical sensor value relating to the air supply to the fuel cell system 10, which is determined using the air mass sensor 22. The oxygen consumption in the stack arrangement 11 and other losses in the fuel cell system 10 are subtracted from this. The physical sensor value of the air mass sensor 22 is output as a mass flow as follows: volf(air_sensor) = mf(physical sensor value) / M(air), where M(air) is approximately 29 g / mol (molar mass of dry air).The oxygen consumption of the stack assembly 11 can be determined according to volf(stack_consumption) = i(stack) * n(stack cells) / Faraday constant / n(electrons_O2), where i(stack) is an electric current flow through the stack assembly 11 measured with a current sensor, n(stack cells) corresponds to the number of stack cells in series through which the current i(stack) flows, the Faraday constant is 96485.3 A*s / mol, and n(electrons_O2) corresponds to four electrons absorbed per O2 molecule. The air losses volf(air_losses) can occur via compressor air bearings and can be determined from the associated geometry and / or empirical reference measurements.

[0045] In a fifth step S50, the model value volf(H2Est_Purge_Mixing_Area) is calculated based on a delay value of the anode exhaust gas from the anode purge valve 19 according to volf(H2Est_Purge_Mixing_Area) = volf(H2Est_Purge, t-50ms). In the example described, the delay value of 50 ms results from the length of the purge channel from the anode purge valve 19 to the mixing area 21.

[0046] In a sixth step S60, the volume flow of the anode exhaust gas is calculated according to volf(H2Est_Purge) = volf(AnodEst_Purge) * con(H2Est_Anode). Accordingly, the fuel or hydrogen content in the anode exhaust gas is estimated from the volume flow through the anode purge valve 19 and the hydrogen content in the anode. The model value con(H2Est_Anode) corresponds to the fuel concentration estimate to be cyclically corrected.

[0047] In a seventh step S70, the volume flow of the anode exhaust gas through the anode purge valve 19 is calculated according to volf(AnodEst_Purge) = mf(AnodEst_Purge) / M(AnodEst), where the volume flow is determined from the mass flow (mf = MassFlow). Here, M(AnodEst) is the molar mass of the anode gas mixture, where M(AnodEst) = M(H2) * con(H2Est_Anode) + M(N2) * (1 - con(H2Est_Anode)). In this case, for simplification, it is approximately assumed that the non-H2 portion is completely occupied by nitrogen, and con(N2Est) = 1 - con(H2Est_Anode) applies.

[0048] In an eighth step S80, the mass flow of the anode exhaust gas through the anode purge valve 19 is determined based on the formula m˙=poA∗kMmRTo(1+k−12)k+1−2(k−1)

[0049] Calculated. Furthermore, mf(AnodEst_Purge) = p0 * A(eff) * Gasterm * Fcoe applies, where p0 corresponds to the gas pressure upstream of the anode purge valve 19 and thus approximately corresponds to the anode pressure at the outlet of the anode 13 and / or in the anode outlet path 17 upstream of the anode purge valve 19. The gas pressure is determined using the pressure sensor 23. The valve effective area is determined according to A(eff) = A(valve) * degree of opening * Kv, where A(valve) is the valve constriction in m 2 depends on the valve geometry or can be measured by the manufacturer. The degree of opening corresponds to the control of the valve according to 1 == 100% opening, and Kv is the specific valve coefficient specified by the valve manufacturer. The gas-specific term Gasterm takes into account the influence of the flowing gas, and Fcoe corresponds to the flow coefficient of the gases.

[0050] In a ninth step S90, the flow coefficient of the gases is determined as Fcoe using the following formula: (1+k−12)k+1−2(k−1)

[0051] In the formula, the variable k is to be determined as an isentropic exponent such that k is a ratio of the specific heat capacities C p to the heat capacity C v where: k=Cp / CV=(Cp,H2*con(H2Est_Anode)+Cp,N2*(1−con(H2Est_Anode))) / (CV,H2*con(H2Est_Anode)+CV,N2*(1−con(H2Est_Anode)))). For simplification, it is approximately assumed that the non-H2 fraction is completely taken up by nitrogen according to con(N2Est) = 1 - con(H2Est_Anode). The specific heat capacity C p and the heat capacity C v are fixed material properties and can be taken from tables.

[0052] In a tenth step S100, the gas-specific term Gasterm can be determined using the following formula part. kMmRTo

[0053] Here, Gasterm = sqrt(k * M m / R / T0), where k was determined in step S90, M m is equal to M(AnodEst) and was determined in step S70, R is a general gas constant and T0 is the anode gas temperature and can be determined as the temperature of the stack assembly by a temperature sensor of the fuel cell system 10.

[0054] The method is performed cyclically to the timing of the anode purge valve 19 in order to further correct the fuel concentration estimate based on the respective calculations and corrections, to bring it to a value as close as possible to an actual fuel concentration value, or to maintain it there. Steps S10 and S20 to S100 can be performed sequentially or at least partially simultaneously and / or in an alternative order.

[0055] In Fig. Figure 6 shows a characteristic curve diagram to further explain the methods described above. In the upper third, a continuous model value graph 33 shows the calculated fuel concentration model values ​​over time. A dotted sensor value graph 34 shows the determined fuel concentration sensor values ​​over time. A dashed offset correction value graph 35 shows offset correction values. Within the framework of the method described above, the respective model difference values ​​are calculated from the minimum and maximum concentration values ​​(valley to peak) of the fuel concentration model values. Analogously, the sensor difference values ​​are determined based on the actual sensor values. Due to sensor inertia inherent in the measuring principle and, in this case, cross-sensitivity with regard to humidity, the sensor values ​​do not drop completely to zero. These factors orInfluences from the sensor signal are determined and used as an offset correction value for the calculation of the fuel concentration model values.

[0056] In the middle third in Fig. 6, the anode purge valve actuation times are shown in an actuation time graph 36. In the lower third, a continuous estimated value graph 37 shows the estimated and / or corrected fuel concentration values ​​within the anode 13 over time. A dotted test sensor graph 38 shows a reference value. Since the test sensor used is not suitable for liquid water, the sensor signal dips downwards twice in the rear area. The reference value is not available for the model calculation according to the method and is used here only to validate and illustrate the function of the method. As shown in the lower third of the Fig. 6, the fuel concentration estimate is slightly increased at each run of the procedure if the model difference value calculated in the upper part of the image is smaller than the sensor difference value and reduced accordingly otherwise. Fig. Figure 6 shows the determination and correction of the fuel concentration estimate during operation of the fuel cell system 10, in which the fuel cell power is continuously varied during the measurement. This particularly affects the stack exhaust air and thus the dilution degree of the purge flow. For this reason, the difference values ​​decrease noticeably after half the time.

[0057] In the example shown, the fuel concentration estimate is adjusted upwards by 2% if a comparison between a sensor difference value and a model difference value shows that the sensor difference value is greater than the model difference value, and is adjusted downwards by 2% if a comparison between a sensor difference value and a model difference value shows that the sensor difference value is smaller than the model difference value. Fig. 7 shows that, if the starting value is incorrect, the corresponding error is continuously corrected by the method until the fuel concentration model value is again close to the fuel concentration sensor value. If the values ​​are equal, no adjustment or correction takes place in any of the illustrated embodiments.

[0058] The invention allows for further design principles in addition to the embodiments illustrated. This means that the invention should not be considered limited to the embodiments explained with reference to the figures. List of reference symbols 10 Fuel cell system 11 Stacking arrangement 12 Cathode 13 Anode 14 Cathode inlet path 15 Cathode outlet path 16 Anode inlet path 17 Anode outlet path 18 Exhaust path 19 Anode purge valve 20 Fuel sensor 21 Mixed area 22 Air mass sensor 23 Pressure sensor 24 fans 25 water separators 26 Inlet valve 27 Diffuser 30 Computer program product 31 Storage medium 32 Control unit 33 Model value graph 34 Sensor value graph 35 Offset correction value graph 36 Control time graph 37 Estimated value graph 38 Test sensor graph 40 electric motor 50 fuel tank 100 vehicles

Claims

[1] A method for correcting fuel concentration estimates of a fuel concentration in an anode (13) of a fuel cell system (10), wherein the fuel cell system (10) comprises a stack assembly (11) with a cathode (12) and the anode (13), a cathode inlet path (14) for conducting oxidant into the cathode (12), a cathode outlet path (15) for conducting cathode exhaust gas from the cathode (12), an anode inlet path (16) for conducting fuel into the anode (13), an anode outlet path (17) for conducting anode exhaust gas from the anode (13), an exhaust path (18) for conducting the cathode exhaust gas and the anode exhaust gas into the environment of the stack assembly (11), and an anode purge valve (19) for selectively conducting anode exhaust gas from the anode outlet path (17) towards of the exhaust gas path (18), comprising: - determining fuel concentration sensor values ​​in the exhaust gas path (18) using a fuel sensor (20) positioned in the exhaust gas path (18), - calculating fuel concentration model values ​​in the exhaust gas path (18) using fuel concentration estimates, - performing comparisons between the fuel concentration sensor values ​​and the fuel concentration model values, and - Correct the fuel concentration estimates based on the comparisons. [2] Method according to claim 1, characterized by , that - the fuel concentration sensor values ​​are determined at least partially during discharge of anode exhaust gas into the outlet path (18) and / or that the fuel concentration model values ​​are calculated at least partially during discharge of anode exhaust gas into the outlet path (18), - when determining the fuel concentration sensor values, lower sensor values ​​for a closed state of the anode purge valve (19), upper sensor values ​​for an open state of the anode purge valve (19) and sensor difference values ​​are determined between a lower sensor value and an upper sensor value, - when calculating the fuel concentration model values, lower model values ​​for a closed state of the anode purge valve (19), upper model values ​​for an open state of the anode purge valve (19) and model difference values ​​are determined between a lower sensor value and an upper sensor value, - when performing the comparisons, the sensor difference values ​​are compared with the model difference values. [3] Method according to claim 2, characterized by , - that a fuel concentration estimate is corrected upwards if a comparison between a sensor difference value and a model difference value shows that the sensor difference value is greater than the model difference value and / or - that a fuel concentration estimate is corrected downwards if a comparison between a sensor difference value and a model difference value shows that the sensor difference value is smaller than the model difference value. [4] Method according to claim 3, characterized by that the fuel concentration estimate is adjusted upwards and / or downwards by less than 10%. [5] Method according to one of the preceding claims, characterized by that the fuel concentration model values ​​are calculated in a time-correlated manner to the fuel concentration sensor values ​​and are each calculated with at least one predefined time delay value. [6] Method according to one of the preceding claims, characterized by that a respective activation time of the anode purge valve (19), in which the anode purge valve (19) is activated to guide anode exhaust gas from the anode outlet path (17) in the direction of the exhaust gas path (18), is used to calculate the respective fuel concentration model value. [7] A fuel cell system (10) comprising a stack assembly (11) with a cathode (12) and an anode (13), a cathode inlet path (14) for conducting oxidant into the cathode (12), a cathode outlet path (15) for conducting cathode exhaust gas from the cathode (12), an anode inlet path (16) for conducting fuel into the anode (13), an anode outlet path (17) for conducting anode exhaust gas from the anode (13), an exhaust path (18) for conducting the cathode exhaust gas and the anode exhaust gas into the environment of the stack assembly (11), an anode purge valve (19) for selectively conducting anode exhaust gas from the anode outlet path (17) towards the exhaust path (18), and a fuel sensor (20) positioned in the exhaust path (18), wherein the fuel cell system (10) is configured and designed to carry out a method according to one of the preceding claims. [8] Vehicle (100) comprising a fuel cell system (10) according to claim 7 and at least one electric motor (40) for driving the vehicle (100), wherein the fuel cell system (10) is configured to supply power to the at least one electric motor (40). [9] Computer program product (30) comprising instructions which, when the computer program product (30) is executed by a computer, cause the computer to carry out the method according to one of claims 1 to 7 in a vehicle (100) according to claim 8. [10] Computer-readable storage medium (31) having a computer program product (30) according to claim 9 stored thereon.

Images