Fuel cell system and method of operating a fuel cell system

Abstract

The presented invention relates to a method (100) for operating a fuel cell system (200), comprising a first detection step (101) in which a hydrogen concentration supplied to a fuel cell stack (203) of the fuel cell system (200) through an inlet valve (209) of the fuel cell system (200) is detected, a second detection step (103) in which a nitrogen volumetric flow rate flowing through the fuel cell stack (203) is detected, and an adjustment step (105) in which a speed of a recirculation blower (205) of the fuel cell system (200) is adjusted based on the detected hydrogen concentration and the detected nitrogen volumetric flow rate. The present invention further relates to a fuel cell system (200) and a computer program product according to the appended claims.

Description

[Technical Field]

[0001] The presented invention relates to a method of operating a fuel cell system, a fuel cell system, and a computer program product. [Background technology]

[0002] Polymer electrolyte membrane (PEM) fuel cell systems convert hydrogen into electrical energy with oxygen, generating waste heat and water.

[0003] A PEM fuel cell consists of an anode supplied with hydrogen, a cathode supplied with air, and a polymer electrolyte membrane between them, which converts the air and oxygen into electricity, water, and heat. To maximize the voltage, several such fuel cells are usually stacked together to form a fuel cell stack.

[0004] In systems where hydrogen is supplied to the PEM anode, the still hydrogen-rich anode exhaust gas is fed back to the anode inlet together with fresh hydrogen by a gas conveying unit, a technique known as recirculation.

[0005] Air is added to the gas mixture flowing through the fuel cell system by a recirculation blower, which is typically operated at a constant speed. Summary of the Invention

[0006] Within the scope of the presented invention, a method for operating a fuel cell system, a fuel cell system, and a computer program product are presented. Further features and details of the invention will become apparent from the respective dependent claims, the following description and the drawings. In that case, features and details described in the context of the method according to the invention naturally also apply in the context of the fuel cell system according to the invention and the computer program product according to the invention, and vice versa, and are therefore always or can be cross-referenced with respect to the disclosure of the individual aspects of the invention.

[0007] The presented invention is used to efficiently operate a fuel cell system, and in particular to dynamically adapt the rotational speed of a recirculation blower of the fuel cell system to the operating conditions of the fuel cell system.

[0008] Thus, according to a first aspect of the present invention, a method for operating a fuel cell system is provided, the method comprising a first detecting step of detecting a concentration of hydrogen supplied to a fuel cell stack of the fuel cell system through an intake valve of the fuel cell system, a second detecting step of detecting a volumetric flow rate of nitrogen flowing through the fuel cell stack, and an adjusting step of adjusting a rotational speed of a recirculation blower of the fuel cell system based on the detected hydrogen concentration and the detected volumetric flow rate of nitrogen.

[0009] In the context of the presented invention, a recirculation blower of a fuel cell system is to be understood as an air conveying unit that supplies air to the fuel cell system by rotating a fan wheel.

[0010] The proposed invention is based on the principle of dynamically adapting the rotation speed of a recirculation blower of a fuel cell system to the current operating conditions of the fuel cell system, i.e., the state of the gas mixture flowing through the fuel cell system, for example to adjust to a predetermined gas mixture ratio. To this end, the operating conditions of the fuel cell system are determined by detecting the concentration of hydrogen supplied to the fuel cell stack of the fuel cell system through the intake valve of the fuel cell system and the volumetric flow rate of nitrogen flowing through the fuel cell stack. Based on the hydrogen concentration and the volumetric flow rate of nitrogen, the amount of air required for optimal or efficient operation of the fuel cell system can be determined, and the rotation speed of the recirculation blower can be adjusted accordingly.

[0011] It is possible to provide that a lambda value for the fuel cell system is determined based on the detected hydrogen concentration and the detected nitrogen volume flow rate, and that the rotational speed of the recirculation blower is adjusted based on this lambda value.

[0012] The lambda value mathematically indicates the operating state of the fuel cell system according to the following equation (1): TIFF0007792007000001.tif16170In that case, "λ H2 " indicates the lambda value of the fuel cell system, TIFF0007792007000002.tif9170 shows the volumetric flow rate of hydrogen supplied to the fuel cell stack, TIFF0007792007000003.tif10170 shows the volumetric flow rate of hydrogen consumed by the fuel cell stack.

[0013] It is further contemplated that the recirculation blower speed may be adjusted to achieve a predetermined lambda value in the fuel cell system.

[0014] A target amount can be predetermined for adjusting the recirculation blower of the fuel cell system, so that the lambda value λ H2 The recirculation blower speed is increased or decreased until the airflow rate equals the target amount.

[0015] Furthermore, in the second detection step, it is possible to detect the nitrogen volume flow rate based on the difference between the pressure upstream of the fuel cell stack in the flow direction of the nitrogen volume flow rate and the pressure downstream of the fuel cell stack in the flow direction of the nitrogen volume flow rate.

[0016] Based on the pressure difference upstream and downstream of the fuel cell stack of each fuel cell system, the nitrogen volumetric flow rate in the fuel cell stack and the resulting hydrogen volumetric flow rate in the fuel cell stack can be estimated in relation to the known hydrogen concentration supplied to the fuel cell stack.

[0017] It is further contemplated that the hydrogen concentration may be detected at least in part by a machine learning machine, without the use of a physical hydrogen concentration sensor.

[0018] A hydrogen concentration sensor can be omitted by using a machine learning machine configured to detect the hydrogen concentration that is metered or adjusted in the fuel cell stack of the fuel cell system based on the operating parameters of the respective fuel cell system.

[0019] Furthermore, the machine learning machine may be trained using a hydrogen concentration sensor in a training fuel cell system and validated based on hydrogen concentration values ​​detected by the hydrogen concentration sensor, in which case the machine learning machine may receive as input signals at least one operating parameter of the recirculation blower of the training fuel cell system and a state parameter of the electrical state of the fuel cell stack of the training fuel cell system.

[0020] In the context of the presented invention, training of a machine learning machine should be understood as a process of varying the mathematical model underlying the machine learning machine until a predetermined goal is achieved, such as a minimum deviation between the results of the model and measurements of hydrogen concentration in a training fuel cell system correspondingly sensed by a hydrogen concentration sensor contemplated by the present invention. In the context of the presented invention, validation of a machine learning machine should be understood as a process in which initial values ​​sensed by the machine learning machine are checked against measured measurements.

[0021] A machine learning machine, such as an artificial neural network or a support vector machine, is trained under monitored conditions, in particular in laboratory operation, and validated using measured values ​​of hydrogen concentration in a training fuel cell system detected using a hydrogen concentration sensor, so that, for example, the deviation between the hydrogen concentration value in the anode circuit of the training fuel cell system detected by the machine learning machine and the hydrogen concentration measured by the hydrogen concentration sensor is minimized or is less than a predetermined threshold, and the machine learning machine can be used in a target fuel cell system that does not have a hydrogen concentration sensor.

[0022] It may further be contemplated that at least a portion of the machine learning machine includes the data model underlying the machine learning machine.

[0023] In the target fuel cell system, a machine learning machine is used to detect the hydrogen concentration in the anode circuit of the target fuel cell system, so that the recirculation blower of the target fuel cell system can be adjusted or operated depending on the hydrogen concentration detected by the machine learning machine, for example, by the machine learning machine simply transmitting the data model underlying the machine learning machine, in whole or in part, to the target fuel cell system.

[0024] The machine learning machine is trained or has been trained using the hydrogen concentration values ​​detected by the hydrogen concentration sensor to interpret each input value, so that when the machine learning machine is fully trained, the machine learning machine is suitable for operating the target fuel cell system without a hydrogen concentration sensor. This means that the fully trained machine learning machine includes a mathematical model of the relationship between each input value and the resulting hydrogen concentration, and this mathematical model includes all operating states performed during the training step of the training fuel cell system, and can be used for open-loop or closed-loop control of the target fuel cell system without a hydrogen concentration sensor based on this.

[0025] In the tests, it was found that in particular operating parameters of the recirculation blower of the respective fuel cell system, such as the output and / or rotation speed of the recirculation blower, state parameters of the electrical state of the fuel cell stack, such as the voltage and / or current of the fuel cell stack, state parameters of the respective fuel cell system, such as the system pressure, as well as characteristic values ​​of the amount of hydrogen supplied to the fuel cell stack by the intake valve, which can be determined, for example, by the operation of the pump, the current supplied to the intake valve, or a flow sensor, are suitable as input values ​​for the machine learning machine.

[0026] According to a second aspect, the presented invention relates to a fuel cell system comprising a monitoring device configured to detect the concentration of hydrogen supplied to a fuel cell stack of the fuel cell system through an intake valve of the fuel cell system, to detect the volumetric flow rate of nitrogen flowing through the fuel cell stack, and to adjust the rotation speed of a recirculation blower of the fuel cell system using the detected hydrogen concentration and the detected volumetric flow rate of nitrogen.

[0027] In a third aspect, the presented invention relates to a computer program product comprising program code means which, when executed on a computer, configures the computer to perform the steps of possible embodiments of the presented method.

[0028] In the context of the presented invention, a computer or monitoring device should be understood as a processor, microcontroller or any other programmable circuit.

[0029] Other advantages, features and details of the invention will become apparent from the following description in which an embodiment of the invention is described in detail with reference to the drawings, in which the features mentioned in the claims and in the description can each alone or in any combination be essential to the invention. [Brief explanation of the drawings]

[0030] [Figure 1] FIG. 1 is a diagram of one possible embodiment of the presented method. [Figure 2] 1 is a diagram of one possible embodiment of the fuel cell system presented. DETAILED DESCRIPTION OF THE INVENTION

[0031] In FIG. 1, a method 100 of operating a fuel cell system is shown.

[0032] The method 100 includes a first detection step 100 in which the concentration of hydrogen supplied to the fuel cell stack of the fuel cell system through the intake valve of the fuel cell system is detected, a second detection step 103 in which the volumetric flow rate of nitrogen flowing through the fuel cell stack is detected, and an adjustment step 105 in which the rotation speed of the recirculation blower of the fuel cell system is adjusted based on the detected hydrogen concentration and the detected volumetric flow rate of nitrogen.

[0033] 2, a fuel cell system 200 is shown comprising a monitoring device 201, a fuel cell stack 203, a recirculation blower 205, a purge valve 207, and an intake valve 209 for metering fresh hydrogen from a tank and the fuel cell stack 203.

[0034] The fuel cell system 200 optionally includes a water separator 211 , a drain valve 213 , and a jet pump 215 for regulating the pressure in the fuel cell stack 203 .

[0035] The monitoring device 201 is configured to detect the concentration of hydrogen supplied to the fuel cell stack 203 by the intake valve 209, to detect the volumetric flow rate of nitrogen flowing through the fuel cell stack 203, and to adjust the rotational speed of the recirculation blower 205 based on the detected hydrogen concentration and the detected volumetric flow rate of nitrogen. [Explanation of symbols]

[0036] 100 ways 101 First detection step 103 Second Detection Step 105 adjustment steps 200 Fuel Cell System 201 Monitoring equipment 203 Fuel Cell Stack 205 Recirculation Blower 207 Purge valve 209 Intake valve 211 Water separator 213 Drain valve 215 Jet Pump

Claims

1. A method (100) of operating a fuel cell system (200), comprising: The method (100) a first detection step (101) in which the concentration of hydrogen supplied to the fuel cell stack (203) of said fuel cell system (200) through the intake valve (209) of said fuel cell system (200) is detected; a second detection step (103) in which the nitrogen volume flow rate through said fuel cell stack (203) is detected; an adjusting step (105) in which the rotation speed of a recirculation blower (205) of the fuel cell system (200) is adjusted based on the detected hydrogen concentration and the detected nitrogen volume flow rate; In the second detection step (103), the nitrogen volume flow rate is determined based on a difference between a pressure upstream of the fuel cell stack (203) in a flow direction of the nitrogen volume flow rate and a pressure downstream of the fuel cell stack (203) in a flow direction of the nitrogen volume flow rate. method.

2. 2. The method (100) of claim 1, wherein a lambda value of the fuel cell system (200) is determined based on the detected hydrogen concentration and the detected nitrogen volume flow rate, and the rotation speed of the recirculation blower (205) is adjusted based on the lambda value.

3. 3. The method (100) according to claim 1 or 2, characterized in that the speed of the recirculation blower (205) is adjusted to achieve a predetermined lambda value in the fuel cell system (200).

4. 3. The method (100) of claim 1 or 2, wherein the hydrogen concentration is sensed at least in part by a machine learning machine without the use of a physical hydrogen concentration sensor.

5. the machine learning machine is trained using a hydrogen concentration sensor in a training fuel cell system and validated based on hydrogen concentration values ​​detected by the hydrogen concentration sensor; 5. The method (100) of claim 4, wherein the machine learning machine receives as input signals at least one operating parameter of a recirculation blower (205) of the training fuel cell system and a state parameter of the electrical state of the fuel cell stack (203) of the training fuel cell system.

6. 5. The method (100) of claim 4, wherein the at least a portion of the machine learning machine includes a data model underlying the machine learning machine.

7. 6. The method (100) of claim 5, wherein the at least a portion of the machine learning machine includes a data model underlying the machine learning machine.

8. A fuel cell system (200) equipped with a monitoring device (201), The monitoring device (201) a hydrogen concentration supplied to a fuel cell stack (203) of the fuel cell system (200) through an intake valve (209) of the fuel cell system (200), a nitrogen volume flow rate flowing through the fuel cell stack (203), and a recirculation blower (205) of the fuel cell system (200) configured to adjust the rotation speed based on the detected hydrogen concentration and the detected nitrogen volume flow rate; the monitoring device (201) determines the nitrogen volume flow rate based on a difference between a pressure upstream of the fuel cell stack (203) in a flow direction of the nitrogen volume flow rate and a pressure downstream of the fuel cell stack (203) in a flow direction of the nitrogen volume flow rate; Fuel cell system.

9. A computer program product comprising program code means for configuring a computer to perform the steps of the method according to claim 1 or 2 when the computer program product is executed on the computer.

Images