JP4887638B2 - Control device for fuel cell system - Google Patents

Description

  The present invention relates to a fuel cell system, and more particularly to a control device for a fuel cell system that learns current-voltage characteristics of a fuel cell.

  A fuel cell is a power generation device that can continue power generation as long as the supply of fuel gas and oxidant gas continues. However, the electrode catalyst and the electrolyte deteriorate in the process of use, and the current-voltage characteristics and the maximum output decrease.

  As a system for learning the characteristics of the current and voltage of the fuel cell that changes with time in this way, there is a system disclosed in Patent Document 1. This technique derives an approximate expression of the relationship between the current and voltage of the fuel cell, and uses the coefficient of each term of the approximate expression as a learning parameter. Then, a learning parameter value is calculated from a detected value of the current voltage of the fuel cell using a general learning algorithm such as a least square method, thereby learning the relationship between the current and the voltage.

In Patent Document 2, a basic output characteristic of the fuel cell is derived based on the pressure of the fuel gas supplied to the fuel cell and the temperature of the fuel cell, and the fuel is calculated using the basic output characteristic and the internal resistance of the fuel cell. A technique for estimating current-voltage characteristics of a battery is disclosed.
JP 2000-357526 A (page 5, FIG. 2) Japanese Patent Laid-Open No. 2002-231295 (page 6, FIG. 2)

  However, the relationship between the output current and the output voltage of the fuel cell also changes depending on the fluctuation of the gas pressure supplied to the fuel cell. That is, even if the output current of the fuel cell is constant, when the gas pressure changes, the output voltage changes, and the relationship between the output current and the output voltage changes temporarily. Therefore, in such a case, when learning of the characteristics of the current and voltage of the fuel cell is performed, the learning result temporarily fluctuates.

In order to solve the above problems, the present invention provides a fuel cell that generates electricity by electrochemically reacting a fuel gas and an oxidant gas, a fuel supply means for supplying fuel gas to the fuel cell, and an oxidizer for the fuel cell. In a control apparatus for a fuel cell system comprising an oxidant gas supply means for supplying an agent gas, an output current detection means for detecting an output current of the fuel cell, and an output voltage detection means for detecting an output voltage of the fuel cell Current voltage characteristic learning means for learning a current voltage characteristic that is a relationship between the output current and the output voltage, state detection means for detecting a state quantity of the fuel cell that affects the current voltage characteristic, based on the amount of variation of the state quantity, and a learning permission means for permitting or prohibiting the learning by the current-voltage characteristic learning means, the learning permission means, of the oxidant gas in a predetermined time And variation of the pressure variation amount and the fuel gas forces, both when the first is equal to or less than the threshold, the fuel cell system according to subject matter to permit learning of the relationship between the output current and the output voltage It is a control device.

  The learning permission means determines the current-voltage characteristics by the current-voltage characteristics learning means according to the state quantity of the fuel cell that affects the current-voltage characteristics of the fuel cell, for example, the amount of fluctuation in the fuel gas pressure, the oxidant gas pressure, and the output current. Is allowed or prohibited, so if the amount of state fluctuation does not affect the current-voltage characteristics of the fuel cell, learning is permitted, and the amount of state fluctuation affects the current-voltage characteristics of the fuel cell. In some cases, learning can be prohibited.

  According to the present invention, when the fluctuation amount of the state quantity that affects the current-voltage characteristics of the fuel cell exceeds the threshold value, the learning of the current-voltage characteristics is not performed, thereby preventing mislearning and always accurate current. Voltage characteristics can be learned.

  Next, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a schematic configuration diagram illustrating the configuration of a fuel cell system including a first embodiment of a control device for a fuel cell system according to the present invention. In FIG. 1, a fuel cell system 1 includes a fuel cell stack 2 using, for example, a solid polymer electrolyte, a hydrogen tank 3 for storing hydrogen as a fuel gas, a hydrogen pressure control valve 4, an ejector 5, a hydrogen circulation channel 7, hydrogen Purge valve 8, compressor 9 for supplying air as oxidant gas, air supply flow path 10, air pressure control valve 11, air inlet temperature sensor 12, air inlet pressure sensor 13, hydrogen inlet temperature sensor 14, hydrogen inlet pressure sensor 15 A current sensor 18 for detecting the output current of the fuel cell stack 2; a voltage sensor 19 for detecting the output voltage of the fuel cell stack 2; a power control device 20; A controller 21 also serving as a permission unit is provided.

  Hydrogen supplied from the hydrogen tank 3 is supplied to the ejector 5 via the hydrogen pressure control valve 4. The hydrogen supplied to the ejector 5 is mixed with the hydrogen that has passed through the hydrogen circulation passage 7 and supplied to the anode of the fuel cell stack 2. The temperature and pressure of hydrogen at the anode inlet of the fuel cell stack 2 are measured by a hydrogen inlet temperature sensor 14 and a hydrogen inlet pressure sensor 15, respectively. The hydrogen pressure control valve 4 is controlled by the controller 21 so that the pressure measured by the hydrogen inlet pressure sensor 15 becomes a pressure corresponding to the required output of the fuel cell. Normally, the hydrogen purge valve 8 is closed, and the hydrogen discharged from the anode outlet of the fuel cell stack 2 circulates to the ejector 5 through the hydrogen circulation passage 7. When water overflow (hereinafter referred to as flooding) or the like occurs in the fuel cell stack 2 or when the operating pressure of the fuel cell stack 2 is reduced, the hydrogen purge valve 8 is opened and the hydrogen circulation channel 7 and the fuel cell stack are opened. The hydrogen present in 2 is discharged.

  Air serving as the oxidant gas is supplied by the compressor 9. The air compressed by the compressor 9 is supplied to the fuel cell stack 2 via the air supply path 10. The pressure of air at the cathode inlet of the fuel cell stack 2 is measured by an air inlet pressure sensor 13 and controlled by an air pressure control valve 11 arranged on the cathode outlet side. The temperature of the air at the cathode inlet of the fuel cell stack 2 is measured by the air inlet temperature sensor 12.

  The air inlet temperature sensor 12, the air inlet pressure sensor 13, the hydrogen inlet temperature sensor 14, the hydrogen inlet pressure sensor 15, the current sensor 18, and the voltage sensor 19 are each connected to an input terminal of the controller 21, and each detection signal is sent to the controller 21. Enter.

  Further, the hydrogen pressure adjustment valve 4, the hydrogen purge valve 8, the compressor 9, the air pressure control valve 11, and the power control device 20 are each connected to an output terminal of the controller 21 and controlled by a control signal output from the controller 21. .

  The controller 21 determines the operating state of the fuel cell stack 2 from the input signals of the sensors, and controls the entire fuel cell system 1. Further, the controller 21 permits or prohibits learning of the current-voltage characteristics of the fuel cell stack 2 based on the fluctuation amount of the state quantity of the fuel cell stack 2 input from each sensor.

  The controller 21 is not particularly limited. In the present embodiment, the controller 21 is constituted by a microprocessor having a CPU, a ROM preliminarily storing control parameters such as a control program and a control map, a working RAM, and an input / output interface. Has been.

  Further, the power taken out from the fuel cell stack 2 is controlled by the power control device 20. This power control device 20 is a step-up / step-down type DC / DC converter capable of stepping up and stepping down an input voltage. The power control device 20 is disposed between the fuel cell stack 2 and an electric load (not shown), and generates power generated by the fuel cell stack 2. Control.

  FIG. 2 is a flowchart for explaining the learning method of the current-voltage characteristics of the fuel cell stack 2 in the first embodiment, which is executed at a predetermined time period. In this embodiment, the hydrogen gas pressure and air pressure supplied to the fuel cell stack are state quantities that affect the current-voltage characteristics of the fuel cell. In the following flowchart, the subscript [t] is added to the current state quantity detection value, and the subscript [t−1] is added to the previous (predetermined time) state quantity detection value.

  First, in step (hereinafter, step is abbreviated as S) 10, the current value It output from the fuel cell stack 2 is detected using the current sensor 18. Next, in S 12, the voltage value Vt output from the fuel cell stack 2 is detected using the voltage sensor 19.

  Next, at S14, the hydrogen inlet pressure sensor 15 is used to detect the hydrogen pressure value PHt at the fuel cell stack 2 inlet, and at S16, the air inlet pressure sensor 13 is used to detect the air pressure value at the fuel cell stack 2 inlet. PAt is detected.

  Next, in S18 and S20, the steady determination of the hydrogen pressure value and the air pressure value is performed. These hydrogen pressure and air pressure cannot learn the current-voltage characteristics of the fuel cell stably because the relationship between the output current and the output voltage varies when the operating pressure varies. Therefore, the change in the current-voltage characteristics of the fuel cell is detected when both the hydrogen pressure and the air pressure are steady.

  Therefore, in S18, when the fluctuation amount | PHt−PHt−1 | of the hydrogen pressure value PHt detected in S14 and the previous value PHt−1 of the hydrogen pressure value is equal to or less than the first threshold value ΔP1, to decide. In S20, when the amount of variation | PAt−PAt−1 | between the air pressure value PAt detected in S18 and the previous value PAt−1 of the air pressure value is equal to or less than the first threshold value ΔP1, it is determined that it is steady.

  If it is not determined to be steady in S18 or S20, the learning of the current-voltage characteristic is prohibited and the process proceeds to S24. If it is determined in S18 and S20 that both the hydrogen pressure value and the air pressure value are steady, learning of the current-voltage characteristics of the fuel cell stack 2 is permitted, and the process proceeds to S22.

  In S22, the current-voltage characteristics of the fuel cell stack 2 are learned using the current-voltage data (It, Vt) detected in S10 and S12.

Here, the current value It and the voltage value Vt are used to represent a linear function that approximates the input (independent variable) as current and the output (dependent variable) as voltage. The linear function is
Y = A · X + B (1)
And formulated. Here, X is current and Y is voltage. Also, as learning parameters, the slope of the current-voltage characteristic is A, and the Y-intercept of the current-voltage characteristic is B.

  The learning value update method uses a sequential least square method for learning parameters based on the error between the measured actual voltage and the learned value obtained by inputting the actual current to the above equation (1). This is achieved by updating with a sequential parameter estimation algorithm.

  As another method for learning current-voltage characteristics, it has table data that takes current as input and outputs voltage as a learning result for the current, and updates the table data based on the relationship between the detected current and voltage. A method of performing learning may be used.

  In S24, the current state quantity (PHt, PAt) of the fuel cell is stored in the storage area of the previous value, and this process ends in preparation for the next execution of this routine.

  According to the present embodiment described above, the oxidant gas pressure and the fuel gas pressure are used as the state quantities affecting the current-voltage characteristics, and the fluctuation amounts of the fuel gas pressure and the oxidant gas pressure are determined to be steady. By setting the threshold value, it is possible to understand the influence of the current-voltage characteristics of the fuel cell due to pressure fluctuations of the oxidant gas and the fuel gas, and to determine whether to allow or prohibit the accurate learning of the current-voltage characteristics. is there.

  As a modification of the present embodiment, the steady determination of the gas pressure uses the target gas pressure of the hydrogen pressure and the air pressure supplied to the fuel cell stack 2, and the fluctuation amount of each target gas pressure is less than a predetermined value. It may be the case. In this case, the target gas pressure is calculated using, for example, the table data shown in FIG. As a result, it is possible to learn the current-voltage characteristics of the fuel cell based on the target gas pressure without detecting the gas pressure supplied to the fuel cell, so that a system that does not affect the noise of the sensor that detects the gas pressure can be configured. .

  Next, Example 2 will be described. The system configuration diagram to which the control device of the fuel cell system of the second embodiment is applied is the same as that of the first embodiment shown in FIG.

  FIG. 4 is a flowchart for explaining a learning method of the current-voltage characteristics of the fuel cell stack 2 in the second embodiment, which is executed at a predetermined time period. The steady judgment of the hydrogen gas pressure and the air pressure is different from the flowchart of the first embodiment shown in FIG. 2, and S32 is provided instead of S30 and S20 instead of S18 of FIG. Since the other steps are the same as those in FIG. 2, the same step numbers are assigned and redundant descriptions are omitted.

  When the hydrogen pressure supplied to the fuel cell stack 2 fluctuates, the relationship between the output current and the output voltage fluctuates, so that the current-voltage characteristics of the fuel cell cannot be learned stably. Therefore, it is necessary to detect the change in the current-voltage characteristics of the fuel cell when the hydrogen pressure is steady.

  For this reason, in S30, when the fluctuation amount | PHt−PHt−1 | of the hydrogen pressure value PHt detected in S14 and the previous value PHt−1 of the hydrogen pressure value is equal to or less than the second threshold value ΔP2, to decide. If the pressure is steady, the process proceeds to S32. If the pressure is not steady, learning is prohibited and the process proceeds to S24.

  Next, in S32, the air pressure value PAt detected in S16 is routinely determined. When the air pressure fluctuates, the relationship between the output current and the output voltage fluctuates, so that the current voltage characteristics of the fuel cell cannot be learned stably. Therefore, it is necessary to detect the change in the current-voltage characteristics of the fuel cell when the air pressure is steady. Here, when the fluctuation amount | PAt−PAt−1 | between the air pressure value PAt detected in S16 and the previous value PAt−1 of the air pressure value is equal to or smaller than the third threshold value ΔP3, it is determined that the air pressure value is constant. Here, since the air pressure has a larger influence on the current-voltage characteristic than the hydrogen pressure, the current-voltage characteristic with higher accuracy can be learned by making the third threshold value ΔP3 smaller than the second threshold value ΔP2. .

  If the pressure is steady, learning of the current voltage characteristics of the fuel cell stack 2 is permitted and the process proceeds to S22. If the pressure is not steady, learning of the current voltage characteristics is prohibited and the process proceeds to S24. The processes in S22 and S24 are the same as those in the first embodiment shown in FIG.

  According to the present embodiment described above, the oxidant gas pressure and the fuel gas pressure are used as the state quantities that affect the current-voltage characteristics of the fuel cell, and the fluctuation amount of the oxidant gas pressure is less than or equal to the second threshold value. If the fuel gas pressure fluctuation amount is less than or equal to the third threshold, learning of the relationship between the output current and the output voltage is permitted, so these thresholds can be set individually. According to the present invention, it is possible to determine whether to permit or prohibit the learning of the current-voltage characteristics of the fuel cell according to the influence of the fluctuation amount of the oxidant gas pressure and the fluctuation amount of the fuel gas pressure on the respective current-voltage characteristics. is there.

  Next, Example 3 will be described. The system configuration diagram to which the control device of the fuel cell system of the third embodiment is applied is the same as that of the first embodiment shown in FIG.

  FIG. 5 is a flowchart for explaining the learning method of the current-voltage characteristics of the fuel cell stack 2 in the third embodiment, which is executed at a predetermined time period. The processes of S40 to S48 are different from the flowchart of the first embodiment shown in FIG. 2, and the processes of S10 to S16 and S22 to S24 are the same.

  In the present embodiment, a determination is made as to whether the load of the fuel cell in S40 is a normal load range or an extremely high load range, and steady determination in the normal load range uses the first threshold value ΔP1 similar to that in the first embodiment. . However, in the steady determination in the extremely high load region of the fuel cell, since the influence of concentration polarization is large in the extremely high load region, the amount of variation in voltage becomes very large with respect to the amount of variation in gas pressure. A seventh threshold value ΔP7 smaller than ΔP1 is used.

  In S40 of FIG. 5, it is determined whether the amount of power generated by the fuel cell is in an extremely high load range. Here, if the output current It of the fuel cell is greater than or equal to the fourth threshold I4, it is determined that the load is extremely high, and the process proceeds to S42. If the output current It is less than the fourth threshold I4, the normal load is Determine and proceed to S46.

  In S42 and S44, it is determined whether or not the amount of fluctuation of the gas pressure in the extremely high load region is steady. Therefore, in S42, the fluctuation amount | PHt−PHt−1 | of the hydrogen pressure value PHt detected in S14 and the previous value PHt−1 of the hydrogen pressure value becomes equal to or smaller than the seventh threshold value ΔP7 (ΔP7 <ΔP1). In this case, it is determined as steady. In S44, when the fluctuation amount | PAt-PAt-1 | between the air pressure value PAt detected in S18 and the previous value PAt-1 of the air pressure value is equal to or less than the seventh threshold value ΔP7, it is determined that the air pressure value is constant.

  If it is not determined to be steady in S42 or S44, learning of the current-voltage characteristic is prohibited, and the process proceeds to S24. If it is determined in S42 and S44 that both the hydrogen pressure value and the air pressure value are steady, learning of the current-voltage characteristics of the fuel cell stack 2 is permitted, and the process proceeds to S22.

  In S46 and S48, it is determined whether or not the fluctuation amount of the gas pressure in the normal load region is steady. Therefore, in S46, when the fluctuation amount | PHt−PHt−1 | of the hydrogen pressure value PHt detected in S14 and the previous value PHt−1 of the hydrogen pressure value is equal to or less than the first threshold value ΔP1, to decide. In S48, when the amount of variation | PAt−PAt−1 | between the air pressure value PAt detected in S18 and the previous value PAt−1 of the air pressure value is equal to or less than the first threshold value ΔP1, it is determined that the air pressure value is constant.

  If it is not determined to be steady in S46 or S48, learning of the current-voltage characteristic is prohibited, and the process proceeds to S24. If it is determined in S46 and S48 that both the hydrogen pressure value and the air pressure value are steady, learning of the current-voltage characteristics of the fuel cell stack 2 is permitted, and the process proceeds to S22. The processes in S22 and S24 are the same as those in the first embodiment shown in FIG.

  According to the present embodiment described above, when the output current of the fuel cell is greater than or equal to the fourth threshold value, the fluctuation amount of the oxidant gas pressure or the fuel gas pressure is determined to be steady as compared with the case where the output current is not the case. Since the first to third threshold values are reduced, the fluctuation amount of the output voltage of the fuel cell with respect to the fluctuation amount of the oxidant gas pressure and the fuel gas pressure in the region where the concentration polarization of the extremely high load of the fuel cell appears remarkably. Therefore, by setting this threshold value, it is possible to determine whether or not the learning of the current-voltage characteristics of the fuel cell is permitted more accurately.

  Next, Example 4 will be described. The system configuration diagram to which the control device of the fuel cell system of the fourth embodiment is applied is the same as that of the first embodiment shown in FIG.

  FIG. 6 is a flowchart for explaining a method for learning the current-voltage characteristics of the fuel cell stack 2 in the fourth embodiment, which is executed at a predetermined time period. In this embodiment, the output current of the fuel cell stack is an amount of state that affects the current-voltage characteristics of the fuel cell.

  First, in S10, the current value It output from the fuel cell stack 2 is detected using the current sensor 18. Next, in S 12, the voltage value Vt output from the fuel cell stack 2 is detected using the voltage sensor 19.

  Next, in S50, it is determined whether the power generation amount of the fuel cell stack 2 is in an extremely low load region or an extremely high load region. Here, when the output current It of the fuel cell is between the fifth threshold value I5 and the sixth threshold value I6 larger than the fifth threshold value I5, it is determined as a normal load and the process proceeds to S52. If it is less than the threshold value I5 or exceeds the sixth threshold value, it is judged as an extremely low load region or an extremely high load region, and the process proceeds to S54.

  Next, in S52, it is determined whether the output current It detected in S10 is steady. This steady state determination is performed in order to remove current voltage data that cannot be stably measured when the load of the fuel cell system changes. Here, when the amount of fluctuation | It−It−1 | between the detected output current It and the previous value It−1 of the output current becomes equal to or smaller than the eighth threshold value ΔI8, it is determined that the current is steady. In addition, as another method for determining the steady state of the power generation state, a method of determining that the current is steady when the variance value of the current measured for a predetermined time is equal to or less than a predetermined value may be applied. If it is determined that the output current is steady, the learning of the current voltage characteristic of the fuel cell is permitted and the process proceeds to S22. If not, the learning of the current voltage characteristic of the fuel cell is prohibited and the process proceeds to S24. move on.

  Next, in S22, the current-voltage characteristics of the fuel cell stack 2 are learned using the current-voltage data (It, Vt) detected in S10 and S12. The learning content of S22 is the same as S22 of the first embodiment.

  In S24, the current state quantity (PHt, PAt) of the fuel cell is stored in the storage area of the previous value, and this process ends in preparation for the next execution of this routine.

  On the other hand, in S54, it is determined whether the output current It detected in S10 in the extremely high load region or the extremely low load region is steady. Since the influence of concentration polarization is large in an extremely high load region, the amount of voltage variation is extremely large with respect to the amount of current variation. Similarly, since the effect of activation polarization is large in the extremely low load region, the amount of voltage fluctuation is very large with respect to the amount of current fluctuation. Therefore, in this region, a ninth threshold value ΔI9 that is smaller than the eighth threshold value ΔI8 is provided, and the fluctuation amount | It−It-1 | between the detected output current It and the previous value It-1 of the output current is the ninth value. Is determined to be steady when the threshold value ΔI9 is less than or equal to. This makes it possible to learn current voltage characteristics with higher accuracy.

  If it is determined that it is steady in S54, learning of the current-voltage characteristics of the fuel cell is permitted and the process proceeds to S22. If not steady, learning of the current-voltage characteristics of the fuel cell is prohibited and the process proceeds to S24. The processes in S22 and S24 are the same as those in the first embodiment shown in FIG.

  As described above, when the output current of the fuel cell is greater than or equal to the fifth threshold and less than or equal to the sixth threshold greater than the fifth threshold, the fluctuation amount of the output current is the eighth threshold. In the case of the following, the learning of the current-voltage characteristics is permitted, and when the output current is less than the fifth threshold value, or when the output current is equal to or more than the sixth threshold value, the fluctuation amount of the output current is Since it is configured to permit learning of current-voltage characteristics when it is less than or equal to the ninth threshold value, which is smaller than the eighth threshold value, an extremely high load concentration polarization region, or an extremely low load activation polarization In the region where the value appears prominently, the amount of fluctuation in the output voltage relative to the variation in the output current of the fuel cell is considered to be larger than in other normal load regions, so this threshold value is compared with the threshold values in other regions. By making it smaller, it meets the characteristics of the fuel cell. Learning of the current-voltage characteristic was able to perform.

  Note that when learning the current-voltage characteristics of the fuel cell, the steady judgment of the gas pressure and the steady judgment of the output current may be performed in combination.

1 is a system configuration diagram illustrating Example 1 of a control device for a fuel cell system according to the present invention. FIG. 3 is a flowchart for explaining learning control processing in the first embodiment. It is an example of table data for calculating a target gas pressure in Example 1. 10 is a flowchart illustrating a learning control process according to the second embodiment. 10 is a flowchart illustrating a learning control process in the third embodiment. 10 is a flowchart illustrating a learning control process according to the fourth embodiment.

Explanation of symbols

1: Fuel cell system 2: Fuel cell stack 3: Hydrogen tank 4: Hydrogen pressure adjustment valve 5: Ejector 7: Hydrogen circulation channel 8: Hydrogen purge valve 9: Compressor 10: Air supply channel 11: Air pressure control valve 12 : Air inlet temperature sensor 13: Air inlet pressure sensor 14: Hydrogen inlet temperature sensor 15: Hydrogen inlet pressure sensor 18: Current sensor 19: Voltage sensor 20: Power control device 21: Controller (current voltage characteristic learning means, learning permission means)

Claims (6)

A fuel cell that generates electricity by electrochemically reacting fuel gas and oxidant gas, fuel supply means for supplying fuel gas to the fuel cell, and oxidant gas supply means for supplying oxidant gas to the fuel cell In the control device of the fuel cell system provided,
Output current detection means for detecting the output current of the fuel cell;
Output voltage detection means for detecting the output voltage of the fuel cell;
Current-voltage characteristic learning means for learning a current-voltage characteristic that is a relationship between the output current and the output voltage;
State detecting means for detecting a state quantity of the fuel cell that affects the current-voltage characteristics;
Learning permission means for permitting or prohibiting learning by the current-voltage characteristic learning means based on the amount of change in the state quantity ,
The learning permission means includes
When the amount of fluctuation in the pressure of the oxidant gas and the amount of fluctuation in the pressure of the fuel gas at a predetermined time are both equal to or less than a first threshold, learning of the relationship between the output current and the output voltage is permitted. A control apparatus for a fuel cell system. The learning permission means includes
When the fluctuation amount of the pressure of the oxidant gas in a predetermined time is less than or equal to the second threshold value and the fluctuation amount of the pressure of the fuel gas is less than or equal to the third threshold value, the output current and the output voltage 2. The fuel cell system control device according to claim 1 , wherein the control device is a means for permitting learning of the relationship. The control device for a fuel cell system according to claim 2 , wherein the second threshold value is set smaller than the third threshold value. If the output current is equal to or greater than the fourth threshold value, the first through instead of the third one of the threshold of claims 1 to, characterized by using a seventh threshold that is set smaller than the respective The control apparatus of the fuel cell system of any one of Claim 3 . Target oxidant gas pressure calculating means for calculating a target pressure of the oxidant gas;
A target fuel gas pressure calculating means for calculating a target pressure of the fuel gas,
The learning permission means includes
On the basis of the amount of fluctuation of the gas pressure of the target amount of change of the target pressure of the oxidant gas and the fuel gas, according to claim 1 or claims and permits learning of the relationship between the output current and the output voltage Item 5. The control device for a fuel cell system according to any one of Items 4 to 5 . The learning permission means includes
When the output current is greater than or equal to a fifth threshold and less than or equal to a sixth threshold greater than a fifth threshold, the amount of variation in the output current is less than or equal to an eighth threshold, Allow learning of the output current and the output voltage,
When the output current is less than the fifth threshold value, or when the output current is greater than or equal to the sixth threshold value, the variation amount of the output current is less than or equal to the ninth threshold value that is less than the eighth threshold value. in some cases, the control device of a fuel cell system according to any one of claims 1 to 5, characterized in that to allow the learning of the output current and the output voltage.

Images