Forward coupling learning for a fuel cell air supply system

Abstract

An air supply system for a fuel cell stack of a fuel cell system includes a compressor, a check valve located in a cathode outlet line, a connection point located in the cathode outlet line connecting an external system to the air supply system, a compressor outlet sensor that determines a measured compressor outlet pressure, a pressure differential request based on a required flow rate measured at the connection point, and one or more controllers in electronic communication with the compressor, the check valve, and the compressor outlet sensor. The one or more controllers learn a parameter of a check valve command via forward feedback when a feedback component of the check valve command is a non-zero value.

Description

INTRODUCTION

[0001] The present disclosure relates to forward coupling learning for controlling a check valve located in a cathode output line of an air supply system for a fuel cell stack.

[0002] A fuel cell system comprises a fuel cell stack, an air supply system, and a fuel supply system. The fuel supply system delivers fuel, such as hydrogen, to the anode of the fuel cell stack, while the air supply system delivers an oxidizer, such as ambient air, to the cathode of the fuel cell stack. The air supply system can further condition the oxidizer before it is delivered to the cathode of the fuel cell stack. More specifically, the oxidizer can be treated by regulating the pressure, temperature, and flow rate before it is received by the cathode of the fuel cell stack.

[0003] Many fuel cell systems may include an expander, typically a turbine, to recover energy from high-pressure exhaust gases released through the cathode of the fuel cell stack. The recovered energy can be used to power other components of the fuel cell system, such as a compressor. Some fuel cell systems may omit the expander and instead include a check valve that generates the pressures required for stable and optimal voltage performance of the fuel cell stack. To achieve comparable transition pressure control performance and diagnostic capability to a fuel cell system using an expander, a control strategy is needed to adjust the check valve in response to the detection of a system deviation.Furthermore, sensors may be required to monitor the pressure at the cathode inlet and cathode outlet of the fuel cell stack, introducing additional components into the fuel cell system.

[0004] Although current fuel cell systems fulfill their intended purpose, there is therefore a need in the field for an improved control strategy for a fuel cell system that includes a check valve. SUMMARY

[0005] An air supply system for a fuel cell stack of a fuel cell system is disclosed according to several aspects. The air supply system includes a compressor arranged along a cathode inlet line of the air supply system, which expels compressed ambient air through the cathode inlet line to a cathode side of the fuel cell stack, where the cathode side of the fuel cell stack generates a cathode exhaust that is sent through a cathode outlet line. The air supply system includes a check valve located in the cathode outlet line, which controls the amount of cathode exhaust supplied to the cathode inlet line. The air supply system includes a connection point located in the cathode outlet line, which connects an external system to the air supply system, wherein the air supply system includes a pressure differential request at a required flow rate measured at the connection point.The air supply system includes a compressor outlet sensor that determines a measured compressor outlet pressure, and one or more controllers in electronic communication with the compressor, the check valve, and the compressor outlet sensor.The one or more controllers contain one or more processors that execute instructions for receiving a request signal and increasing a compressor outlet pressure setpoint in order to achieve, in response to receiving the request signal, conditions required for the fuel cell stack to generate maximum power output, and for determining a check valve command to actuate the check valve to a position that causes the measured compressor outlet pressure to equal the compressor outlet pressure setpoint, whereby a forward-feedback learned parameter of the check valve command is learned when a feedback component of the check valve command is a non-zero value.

[0006] According to another aspect, one or more controllers execute instructions to calculate an estimated setpoint-based inlet pressure and an estimated setpoint-based outlet pressure of the check valve, which are required to achieve the compressor outlet pressure setpoint.

[0007] According to yet another aspect, the estimated inlet pressure based on the target value is determined on the basis of the following: pBPVINest(SP)=psp−dVSp⋅klam−dVSp2⋅(kturb+klearn), where dV Sp represents a setpoint for the volumetric airflow rate of the air supply system, k lam a parameter for laminar flow, k turb a parameter for turbulent flow, k learn The parameter learned through feedforward feedback, which is initially set to 0, is p BPVINest(SP) represents the estimated inlet pressure based on the target value and psp represents the compressor outlet pressure setpoint.

[0008] According to one aspect, the estimated outlet pressure based on the target value is determined on the basis of the following: pBPVOutest(SP)=pamb+X, where P SPVOUTest(SP) X represents the estimated outlet pressure based on the setpoint, X represents the pressure differential requirement at the required flow rate, and p amb represents the ambient pressure.

[0009] According to another aspect, one or more controllers execute instructions to calculate a forward coupling component based on the flow resistance as a function of a characteristic position of the check valve, which is based on the estimated inlet pressure of the check valve based on the setpoint and on the estimated outlet pressure of the check valve based on the setpoint.

[0010] According to yet another aspect, the forward coupling component of the check valve command, which is based on the flow resistance as a function of a characteristic position of the check valve, is determined by the following: FFWDkV=4.633⋅flowspMW⋅TpBPVINest(SP)2−pBPVOUTest(SP)2, where FFWD kV the forward coupling component of the check valve command, which is based on the flow resistance as a function of a characteristic position of the check valve, p BPVINest(SP) P represents the estimated inlet pressure based on the target value. SPVOUTest(SP) MW represents the estimated outlet pressure based on the setpoint, MW represents the molecular weight of the air, and T represents the temperature of the air.

[0011] According to one aspect, the one or more controllers execute instructions to determine a current value for an exhaust gas deviation value, where the current value for the exhaust gas deviation value is the difference between the compressor outlet pressure measured by the compressor outlet sensor and the compressor outlet pressure setpoint.

[0012] According to one aspect, the one or more controllers execute instructions to generate the check valve command, which actuates the check valve into a position that meets the differential pressure requirement.

[0013] According to yet another aspect, one or more controllers execute instructions to determine a feedback component of the check valve command, wherein the feedback component is a difference between the check valve command and the forward feedback component of the check valve command.

[0014] According to one aspect, the one or more controllers execute instructions to calculate an estimated feedback-based inlet pressure and an estimated feedback-based outlet pressure of the check valve based on a pressure drop model of the check valve and a check valve command based on the flow resistance as a function of a characteristic position of the check valve.

[0015] According to another aspect, the estimated feedback-based inlet pressure and the estimated feedback-based outlet pressure of the check valve are determined based on the following: pBPVINest(FB)=pBPVOUTest(FB)2+4.6332⋅flowsp2⋅MW⋅T(PoscmdkV)2, where p BPVINest(FB) the estimated inlet pressure based on feedback, P BPVOUTest(FB)The estimated feedback-based outlet pressure is represented, MW represents the molecular weight of the air, T represents the temperature of the air, and Pos cmdkV represents the check valve command, which is based on the flow resistance as a function of a characteristic position of the check valve.

[0016] According to another aspect, one or more controllers, based on a feedforward-feedback physics model, the estimated feedback-based inlet pressure, and the estimated feedback-based outlet pressure, execute instructions to resolve for the parameter learned through feedforward feedback.

[0017] According to yet another aspect, the parameter learned through feedforward feedback is determined based on the following: klearn=psp−pBPVINest(FB)−dVSp⋅klamdVSp2−kturb, where dV Sprepresents a setpoint for the volumetric airflow rate of the air supply system, k lam represents a parameter of the laminar flow, k turb represents a parameter of the turbulent flow, k learn The parameter learned through feedforward feedback, which is initially set to 0, is P BPVINest(SP) represents the estimated inlet pressure based on feedback and p sp represents the compressor outlet pressure setpoint.

[0018] According to one aspect, one or more controllers store the parameter learned through forward feedback in non-volatile memory.

[0019] According to another aspect, one or more controllers execute instructions to reset an integral gain of one or more controllers when the parameter learned by feedforward feedback is saved in non-volatile memory.

[0020] According to one aspect, an air supply system for a fuel cell stack of a fuel cell system is disclosed. The air supply system includes a compressor arranged along a cathode inlet line of the air supply system, which expels compressed ambient air through the cathode inlet line to a cathode side of the fuel cell stack, where the cathode side of the fuel cell stack generates a cathode exhaust that is sent through a cathode outlet line. The air supply system includes a check valve located in the cathode outlet line, which controls the amount of cathode exhaust supplied to the cathode inlet line. The air supply system includes a connection point located in the cathode outlet line, which connects an external system to the air supply system, wherein the air supply system includes a pressure differential request at a required flow rate measured at the connection point.The air supply system includes a compressor outlet sensor that determines a measured compressor outlet pressure and one or more controllers in electronic communication with the compressor, the check valve, and the compressor outlet sensor. The one or more controllers contain one or more processors that execute instructions to receive a request signal and increase a compressor outlet pressure setpoint in order to achieve, in response to the received request signal, the conditions required for the fuel cell stack to generate maximum power output.One or more controllers determine a check valve command to actuate the check valve to a position that causes the measured compressor outlet pressure to equal the compressor outlet pressure setpoint. A forward-learned parameter of the check valve command is learned when a feedback component of the check valve command is a non-zero value. The one or more controllers store the forward-learned parameter in non-volatile memory and reset an integral gain of the one or more controllers when the forward-learned parameter is stored in non-volatile memory.

[0021] According to another aspect, one or more controllers execute instructions to resolve for the parameter learned through forward feedback, based on a feedforward-feedback physics model, an estimated feedback-based inlet pressure, and an estimated feedback-based outlet pressure.

[0022] According to another aspect, the parameter learned through feedforward feedback is determined based on the following: klearn=psp−pBPVINest(FB)−dVSp⋅klamdVSp2−kturb, where dV Sp represents a setpoint for the volumetric airflow rate of the air supply system, k lam represents a parameter of the laminar flow, k turb represents a parameter of the turbulent flow, k learn The parameter learned through feedforward feedback, which is initially set to 0, is P BPVINest(SP)represents the estimated inlet pressure based on feedback and p sp represents the compressor outlet pressure setpoint.

[0023] According to yet another aspect, one or more controllers execute instructions to calculate an estimated setpoint-based inlet pressure and an estimated setpoint-based outlet pressure of the check valve, which are necessary to achieve the compressor outlet pressure setpoint.

[0024] According to one aspect, an air supply system for a fuel cell stack of a fuel cell system that is part of a vehicle is disclosed. The air supply system includes a compressor arranged along a cathode inlet line of the air supply system, which expels compressed ambient air through the cathode inlet line to a cathode side of the fuel cell stack, where the cathode side of the fuel cell stack generates a cathode exhaust that is sent through a cathode outlet line. The air supply system includes a check valve located in the cathode outlet line that controls the amount of cathode exhaust supplied to the cathode inlet line.The air supply system includes a connection point located in the cathode outlet line, which connects an external system to the air supply system. The air supply system contains a differential pressure request at a requested flow rate measured at the connection point. The air supply system includes a compressor outlet sensor that determines a measured compressor outlet pressure and one or more controllers in electronic communication with the compressor, the check valve, and the compressor outlet sensor. The one or more controllers contain one or more processors that execute instructions to receive a request signal and increase a compressor outlet pressure setpoint in order to achieve, in response to the received request signal, conditions necessary for the fuel cell stack to generate maximum power output.One or more controllers determine a check valve command to actuate the check valve to a position that causes the measured compressor outlet pressure to equal the compressor outlet pressure setpoint. A forward-feedback learned parameter of the check valve command is learned when a feedback component of the check valve command is a non-zero value. Based on a forward-feedback physics model, the estimated feedback-based inlet pressure, and the estimated feedback-based outlet pressure, the one or more controllers resolve for the forward-feedback learned parameter, which is determined based on the following: klearn=psp−pBPVINest(FB)−dVSp⋅klamdVSp2−kturb, where dV Sprepresents a setpoint for the volumetric airflow rate of the air supply system, k lam represents a parameter of the laminar flow, k turb represents a parameter of the turbulent flow, k learn The parameter learned through feedforward feedback, which is initially set to 0, is p BPVINest(SP) represents the estimated inlet pressure based on feedback and p sp This represents the compressor outlet pressure setpoint. One or more controllers store the parameter learned through feedforward in non-volatile memory and reset an integral gain of one or more controllers when the parameter learned through feedforward is stored in non-volatile memory.

[0025] Further areas of applicability are evident from the description given here. It should be understood that the description and the specific examples are for illustrative purposes only and are not intended to limit the scope of protection afforded by this disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings described here serve only for illustration and are not intended to limit the scope of protection of the present disclosure in any way; they show: Fig. 1 a schematic block diagram of a vehicle comprising a fuel cell system having one or more controllers in electronic communication with a check valve, according to an exemplary embodiment; and Fig. 2 a schematic representation of the software architecture of one or more in Fig. 1 controller shown according to an exemplary embodiment. DETAILED DESCRIPTION

[0027] The following description is by its very nature merely exemplary and is not intended to limit the present disclosure, application or uses.

[0028] Based on Fig. Figure 1 contains a schematic block diagram of an exemplary vehicle 8, as shown, comprising a fuel cell system 10, which includes a fuel cell stack 12. It should be noted that the vehicle 8 can be any type of vehicle, such as a sedan, truck, off-road vehicle, van, or motorhome, but is not limited to these. The fuel cell system 10 includes an air supply system 14 and a fuel supply system 16. The air supply system 14 includes an ambient air source 20, a compressor 22, a compressor outlet sensor 24, a water vapor transfer unit 26, a check valve 28, a connection point 32, a cathode inlet line 34, a cathode outlet line 36, and one or more controllers 40. The connection point 32 is located in the cathode outlet line 36, which connects an external system to the air supply system 14 of the fuel cell system 10.The connection point 32 can be directly connected to the ambient air source 20 or to a specific exhaust duct system. The check valve 28 is located in the cathode outlet line 36 of the air supply system 14, upstream of the connection point 32. The one or more controllers 40 are in electronic communication with the compressor 22, with the compressor outlet sensor 24, and with the check valve 28. According to embodiments, the one or more controllers 40 represent one or more integral gain controllers. The compressor outlet sensor 24 is located at an outlet 38 of the compressor 22 and determines a measured compressor outlet pressure p. outlet .

[0029] The fuel supply system 16 includes a fuel source 42, an injection device 44, and an anode inlet line 46. The fuel source 42 contains a high-pressure reservoir that holds a fuel source, such as hydrogen fuel or gas, which is supplied to the injection device 44. The injection device 44 injects a controlled quantity of fuel into an anode side 52 of the fuel cell stack 12 via the anode inlet line 46.

[0030] The compressor 22 of the air supply system 14 is located in the cathode inlet line 34 and receives ambient air from the ambient air source 20. It then expels compressed ambient air via the cathode inlet line 34 through the water vapor transfer unit 26 to a cathode side 58 of the fuel cell stack 12. The cathode side 58 of the fuel cell stack 12 generates a cathode exhaust gas, which is sent through the cathode outlet line 36. The check valve 28 is located upstream of the connection point 32 in the cathode outlet line 36 and controls the pressure in the air supply system 14. As explained in more detail below, the one or more controllers 40 determine a check valve command P. OSCMD , which instructs the check valve 28 to modulate the pressure at the outlet 38 of the compressor 22.

[0031] It should be noted that the air supply system 14 includes a pressure differential requirement X at a required flow rate, measured at connection point 32. By way of example only, the pressure differential requirement X according to one embodiment at the required flow rate is 8 kilopascals (kPa) at 150 grams per second (g / s), although it should be noted that other values ​​can also be used. It should be noted that the actual pressure differential measured at connection point 32 during operation of the fuel cell system 10 may deviate from the pressure differential requirement X at the required flow rate by an exhaust gas deviation value Y. In other words, the exhaust gas deviation value Y represents the difference between the pressure differential requirement X at the required flow rate and the actual pressure differential at connection point 32.

[0032] If the actual pressure difference measured at connection point 32 deviates from the pressure difference requirement X at the required flow rate by the exhaust gas deviation value Y, one or more controllers 40 can send a user-generated request signal 60 (in Fig. (2 shown) to increase the power output of the fuel cell stack 12 to its maximum power output. The request signal 60 is directly mapped to an absolute pressure setpoint at the outlet 38 of the compressor 22. Thus, in response to the user-generated request 60, one or more controllers 40 increase a compressor outlet pressure setpoint p. sp , in order to achieve the conditions required for the fuel cell stack 12 to generate the maximum power output requested by the user. The compressor outlet pressure setpoint p spThis can also be referred to as the fuel system pressure setpoint and represents the pressure setpoint at the compressor outlet 38 22. It should be noted that the one or more controllers 40 instruct the flow through the fuel cell stack 12 to increase while the compressor outlet pressure setpoint p sp increases. The conditions required for the fuel cell stack 12 to generate maximum power output are based on a relationship between the compressor outlet pressure setpoint p. sp and the absolute pressure setpoint at the outlet 38 of the compressor 22. If the absolute pressure setpoint at the outlet 38 of the compressor 22 does not correspond to the compressor outlet pressure setpoint p sp If this were increased, the voltage across the fuel cell stack 12 would be suboptimal at medium to high current setpoints.

[0033] As explained below, one or more controllers 40 learn a parameter k learned through feedforward feedback. learn of the check valve command P OSCMD , if a feedback component FB of the check valve P OSCMD is a non-zero value. The check valve command P OSCMD actuates the check valve 28 into a position that causes the compressor outlet pressure p, determined by the compressor outlet sensor 24, to be applied. outlet equal to the compressor outlet pressure setpoint p sp is. The check valve command P OSCMD is the sum of a forward feedback component FFWD and the feedback component FB of the check valve command P OSCMD , or P OSCMD = FFWD + FB. The feedback component FB of the check valve command P OSCMDrepresents an uncertainty estimate, while one or more controllers 40 use the parameter k learned through feedforward feedback learn of the check valve command P OSCMD learn.

[0034] Fig. 2 is a representation of the software architecture of one or more in Fig. 1 controller 40 shown. Based on both of Fig. 1 as well as from Fig. 2. The one or more controllers 40 contain a pressure setpoint block 70, a forward feedback block 72, a pressure error block 74, a position command block 76, a feedback block 78, a back pressure calculation block 80, a learned value block 82, and a memory block 84. The pressure setpoint block 70 of the one or more controllers 40 receives the request signal 60, which indicates that the one or more controllers 40 should set the compressor outlet pressure setpoint p. spincrease to achieve conditions required for the fuel cell stack 12 to generate maximum power output. It should be noted that the one or more controllers 40 may include a user interface (not shown), such as a keyboard or touchscreen, which allows a user to input the user-generated request signal 60. In response to receiving the request signal 60, the pressure setpoint block 70 of the one or more controllers 40 increases the compressor outlet pressure setpoint p. sp , in order to achieve the conditions required for the fuel cell stack 12 to generate maximum power output.

[0035] The pressure setpoint block 70 sends the compressor outlet pressure setpoint p spto the forward feedback block 72 of one or more controllers 40. The forward feedback block 72 calculates an estimated inlet pressure p based on the setpoint, using a model based on forward feedback physics. BPVINest(SP) and an estimated outlet pressure P based on the setpoint SPVOUTest(SP) of the check valve 28, which are required to achieve the compressor outlet pressure setpoint p calculated by the pressure setpoint block 70 sp to achieve this. According to one embodiment, the model based on feedforward physics calculates the estimated inlet pressure p based on the setpoint. BPVINest(SP) of the check valve 28 based on equation 1 and the estimated outlet pressure P based on the setpoint SPVOUTest(SP) based on equation 2, which are as follows: pBPVINest(SP)=psp−dVSp⋅klam−dVSp2⋅(kturb+klearn) PBPVOutest(SP)=pamb+X, where dV Sprepresents a setpoint of the volumetric air flow rate of the air supply system 14, k lam a parameter for laminar flow, which is based on the required airflow as a function of the pressure drop properties of the air supply system 14, represents k turb a parameter for turbulent flow, which is based on the required airflow as a function of pressure drop properties for the air supply system 14, represents k learn the parameter learned through forward feedback, which is initially set to 0, is and p amb represents the ambient pressure. It should be noted that the parameter k lam for laminar flow and the parameter k turb For turbulent flow, these are static values ​​that do not change.

[0036] Furthermore, the forward coupling block 72 calculates, based on the flow resistance as a function of a characteristic position of the check valve 28, the estimated inlet pressure p based on the setpoint. BPVINest(SP) of the check valve 28 and the estimated outlet pressure P based on the setpoint SPVOUTest(SP) of the check valve 28 a forward coupling component FFWD kV a check valve command Pos cmdkV based on equation 3, which is as follows: FFWDkV=4.633⋅flowspMW⋅TpBPVINest(SP)2−pBPVOUTest(SP)2, where flow sp MW represents the setpoint of the molar air flow rate, MW represents the molecular weight of the air, and T represents the temperature of the air.

[0037] It should be noted that the differential pressure requirement X is not met at the required flow rate if the forward coupling block 72 of one or more controllers 40 actuates the position of the check valve 28, resulting in a pressure higher than the compressor outlet pressure setpoint p. sp is, leads. Thus, the pressure error block 74 of one or more controllers 40 determines a current value for the exhaust gas deviation value Y, which is the difference between the compressor outlet pressure p determined by the compressor outlet sensor 24. outlet and the compressor outlet pressure setpoint p sp , or p outlet - p sp , where the compressor outlet pressure setpoint p spis modeled. In the case of an internal pressure differential deviation from the pressure differential requirement X at the required flow rate in the air supply system 14, the exhaust gas deviation value Y can represent an uncertainty estimate at the fuel cell stack 12 in the pressure at the cathode inlet line 34.

[0038] The position command block 76 receives the current value for the exhaust gas deviation value Y from the pressure error block 74 of one or more controllers 40. Based on the current value for the exhaust gas deviation value Y, the position command block 76 generates the check valve command P. OSCMD , which actuates the check valve 28 into a position to set the compressor outlet pressure setpoint p sp to fulfill. More precisely, in response to the determination that the current value for the exhaust gas deviation value Y is positive, the position command block 76 issues the check valve command P. OSCMDto further open the check valve 28 in order to reduce the pressure. In response to the determination that the current value for the exhaust gas deviation value Y is negative, the position command block 76 issues the check valve command P. OSCMD to further close the check valve 28 in order to increase the pressure.

[0039] Subsequently, the feedback block 78 of one or more controllers 40 determines the feedback component FB of the check valve command P. OSCMD More precisely, the feedback component FB is the difference between the check valve command P OSCMD and the forward feedback component FFWD, or FB = P OSCMD - FFWD. As mentioned above, the feedback component FB of the check valve command P represents OSCMDthe uncertainty estimate. The feedback block 78 of one or more controllers 40 can also compare the exhaust gas deviation value Y with an exhaust gas deviation threshold. In response to the determination that the exhaust gas deviation value Y exceeds the exhaust gas deviation threshold, the feedback block 78 can instruct the air supply system 14 to cease operation. The exhaust gas deviation threshold is determined based on the operating limits of the air supply system 14.

[0040] Subsequently, the back pressure calculation block 80 of one or more controllers 40 calculates an estimated feedback-based inlet pressure p based on a pressure drop model of the check valve 28. BPVINest(FB) and an estimated, feedback-based outlet pressure P BPVOUTest(FB)of the check valve 28. According to one embodiment, the pressure drop model of the check valve 28 determines the estimated feedback-based inlet pressure p. BPVINest(FB) based on the estimated feedback-based outlet pressure p BPVOUTest(FB) and the check valve command position cmdkV based on a flow resistance depending on a characteristic position of the check valve 28 based on equation 4, which is as follows: pBPVINest(FB)=pBPVOUTest(FB)2+4.6332⋅flowsp2⋅MW⋅T(PoscmdkV)2.

[0041] Subsequently, block 82 of the learned value of one or more controllers 40 solves, based on the forward-feedback physics model, the estimated feedback-based inlet pressure p. BPVINest(FB) and the estimated feedback-based outlet pressure p BPVOUTest(FB) after the parameter k learned through feedforward feedback learnAccording to one embodiment, the feedforward-feedback physics-based model calculates the parameter k learned through feedforward feedback. learn based on equation 5, which is derived from equation 1 and is as follows: klearn=psp−pBPVINest(FB)−dVSp⋅klamdVSp2−kturb.

[0042] Subsequently, memory block 84 receives the parameter k learned by feedforward from block 82 of the learned value of one or more controllers 40. learn and secures the parameter k learned through forward feedback learn in the non-volatile memory of one or more controllers 40. Furthermore, memory block 84 of one or more controllers 40 resets an integral gain of one or more controllers 40 when the parameter k learned by feedforward feedback learnis stored in non-volatile memory to prevent the value of the check valve command P from changing. OSCMD changes.

[0043] In general, based on the figures, the disclosed air supply system for a fuel cell system provides various technical effects and advantages. More specifically, the forward-feed learning algorithm allows the air supply system to include a check valve in the cathode output line instead of an expander. The forward-feed learning algorithm determines a valve command that instructs the check valve to actuate in response to the determination that the actual pressure differential measured at the connection point does not meet the pressure differential requirement at the required flow rate. Furthermore, the forward-feed learning algorithm also determines a feedback component of the valve command, where the feedback component represents an uncertainty estimate.Finally, the forward-feedback learning algorithm eliminates the need for pressure sensors at the cathode inlet and cathode outlet of the fuel cell stack, thus reducing the complexity of the fuel cell system.

[0044] Controllers can refer to, or be part of, an electronic circuit, a combination logic circuit, a free programmable logic array (FPGA), a processor (shared, dedicated, or grouped) that executes code, or a combination of some or all of the above, such as in a system-on-a-chip. Additionally, controllers can be microprocessor-based, such as a computer with at least one processor, memory (RAM and / or ROM), and associated input and output buses. The processor can operate according to the control of an operating system residing in memory. The operating system can manage computer resources in such a way that computer program code, embodied as one or more computer software applications, such as an application residing in memory, can contain instructions to be executed by the processor.According to an alternative embodiment, the processor can execute the application directly, and the operating system can be omitted in this case.

[0045] The description in this disclosure is essentially only exemplary, and modifications that do not deviate from the main point of this disclosure shall remain within the scope of protection of this disclosure. Such modifications shall not be considered a deviation from the inventive concept and the scope of protection of this disclosure.

Claims

[1] Air supply system for a fuel cell stack of a fuel cell system, wherein the air supply system comprises: a compressor arranged along a cathode inlet line of the air supply system, which expels compressed ambient air via the cathode inlet line to a cathode side of the fuel cell stack, where the cathode side of the fuel cell stack generates a cathode exhaust which is sent through a cathode outlet line; a check valve located in the cathode outlet line that controls the amount of cathode exhaust supplied to the cathode inlet line; a connection point located in the cathode outlet line connecting an external system to the air supply system, wherein the air supply system includes a differential pressure requirement at a required flow rate measured at the connection point; a compressor outlet sensor that determines a measured compressor outlet pressure; and one or more controllers in electronic communication with the compressor, with the check valve and with the compressor outlet sensor, wherein the one or more controllers contain one or more processors that execute instructions to: Receiving a request signal; Increasing a compressor outlet pressure setpoint to achieve, in response to the receipt of the request signal, conditions required for the fuel cell stack to generate maximum power output; and Determining a check valve command to actuate the check valve to a position that causes the measured compressor outlet pressure to be equal to the compressor outlet pressure setpoint, wherein a forward-feedback learned parameter of the check valve command is learned when a feedback component of the check valve command is a non-zero value. [2] Air supply system according to claim 1, wherein one or more controllers execute instructions to: Calculating an estimated setpoint-based inlet pressure and an estimated setpoint-based outlet pressure of the check valve required to achieve the compressor outlet pressure setpoint. [3] Air supply system according to claim 2, wherein the estimated inlet pressure based on the setpoint is determined on the basis of the following: pBPVINest(SP)=psp−dVSp⋅klam−dVSp2⋅(kturb+klearn), where dVSp represents a setpoint for the volumetric airflow rate of the air supply system, k lam a parameter for laminar flow, k turb a parameter for turbulent flow, k learn The parameter learned through feedforward feedback, which is initially set to 0, is p BPVINest(SP) represents the estimated inlet pressure based on the target value and p sp represents the compressor outlet pressure setpoint. [4] Air supply system according to claim 3, wherein the estimated outlet pressure based on the setpoint is determined on the basis of the following: PBPVOutest(SP)=pamb+X, where P BPVOUTest(sp) X represents the estimated outlet pressure based on the setpoint, X represents the pressure differential requirement at the required flow rate, and p amb represents the ambient pressure. [5] Air supply system according to claim 2, wherein the one or more controllers execute instructions to: Calculating a forward coupling component based on the flow resistance as a function of a characteristic position of the check valve, which is based on the estimated inlet pressure of the check valve based on the setpoint and on the estimated outlet pressure of the check valve based on the setpoint. [6] Air supply system according to claim 5, wherein the forward coupling component of the check valve command, which is based on the flow resistance as a function of a characteristic position of the check valve, is determined by the following: FFWDkV=4.633⋅flowspMW⋅TpBPVINest(SP)2−pBPVOUTest(SP)2, where FFW D kVThe forward coupling component of the check valve command is represented, which is based on the flow resistance as a function of a characteristic position of the check valve, P BPVINest(SP) P represents the estimated inlet pressure based on the target value. SPVOUTest(SP) MW represents the estimated outlet pressure based on the setpoint, MW represents the molecular weight of the air, and T represents the temperature of the air. [7] Air supply system according to claim 5, wherein the one or more controllers execute instructions to: Determining a current value for an exhaust gas deviation value, where the current value for the exhaust gas deviation value is the difference between the compressor outlet pressure measured by the compressor outlet sensor and the compressor outlet pressure setpoint. [8] Air supply system according to claim 7, wherein one or more controllers execute instructions to: Generating the check valve command that actuates the check valve to a position that meets the differential pressure requirement. [9] Air supply system according to claim 8, wherein the one or more controllers execute instructions to: Determining a feedback component of the check valve command, wherein the feedback component is a difference between the check valve command and the forward feedback component of the check valve command. [10] Air supply system according to claim 9, wherein the one or more controllers execute instructions to: Calculating an estimated feedback-based inlet pressure and an estimated feedback-based outlet pressure of the check valve based on a pressure drop model of the check valve and a check valve command based on the flow resistance as a function of a characteristic position of the check valve.

Images