JP2006244966A - Current-voltage learning device of fuel cell - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a current-voltage learning device of a fuel cell that can prevent the error of the learning of the current-voltage character by stopping the learning according to parameters of the condition of supply gas or the condition of the humidity of the fuel cell. <P>SOLUTION: A I-V learning system 7 learns the current voltage character of the fuel cell 1 based on a current detected by a current sensor 5 and a voltage detected by a voltage sensor 6. A learning stopping system 8 judges the learning of the I-V learning system 7 whether stop or not according to an air flow rate standard value fixed according to the target take off current of the fuel cell and an air flow rate detected by an air flow sensor 3. When it is judged to stop, the learning is stopped by sending a stopping signal to the I-V learning system 7. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an apparatus for learning current-voltage characteristics of a fuel cell.

  In a fuel cell, a fuel gas such as hydrogen gas and an oxidizing gas containing oxygen are electrochemically reacted through an electrolyte, and electric energy is directly taken out between electrodes provided on both surfaces of the electrolyte. In particular, a polymer electrolyte fuel cell using a polymer electrolyte has attracted attention as a power source for electric vehicles because of its low operating temperature and easy handling. That is, a fuel cell vehicle is equipped with a hydrogen storage device such as a high-pressure hydrogen tank, a liquid hydrogen tank, or a hydrogen storage alloy tank in the vehicle, and reacts by supplying hydrogen supplied therefrom and air containing oxygen to the fuel cell. This is the ultimate clean vehicle that drives the motor connected to the drive wheels with the electric energy extracted from the fuel cell, and the only exhaust material is water.

  A fuel cell is a power generator that can continue power generation as long as the supply of fuel gas and oxidant gas continues. However, in the process of use, components such as the electrode catalyst and the electrolyte membrane are deteriorated, and the power generation performance is lowered.

As a system for diagnosing deterioration of a fuel cell based on the current-voltage characteristics of the fuel cell, for example, there is a deterioration diagnosis method disclosed in Patent Document 1. This method detects the output current and output voltage of the fuel cell, learns the current-voltage characteristic of the fuel cell based on the output current and output voltage, and based on the slope of the current-voltage characteristic and the Y-intercept, This is a method for diagnosing deterioration of a fuel cell.
JP-A-11-195423 (page 6, FIG. 1)

  However, in the above conventional example, when the flow rate of the gas supplied to the fuel cell is insufficient or the humidification state deteriorates, the voltage drops even though the fuel cell is not deteriorated, and the current-voltage characteristics are improved. There was a problem of mislearning. In addition, when the fuel cell is clogged with water, the cell voltage is lowered, which causes a problem of erroneous learning.

  In order to solve the above problems, the present invention learns the current-voltage characteristics of a fuel cell that generates electricity by an electrochemical reaction between a fuel gas and an oxidant gas through an electrolyte membrane sandwiched between an anode and a cathode. In a current-voltage characteristic learning device for a fuel cell, current-voltage characteristic learning means for inputting the current value and voltage value of the fuel cell to learn the current-voltage characteristic of the fuel cell, fuel gas supplied to the fuel cell, oxidation And a learning stop means for stopping the learning of the current-voltage characteristic learning means when the fluid state of any one of the agent gas, the coolant, and the humidifying water deviates from a desired operation state. .

  According to the present invention, when the state of any one of the fuel gas, oxidant gas, coolant, and humidifying water supplied to the fuel cell deviates from a desired operation state, the current-voltage characteristic learning means learns. When the supply gas flow rate to the fuel cell is insufficient or the humidification state deteriorates, the voltage decreases even though the fuel cell is not deteriorated. By stopping the learning of characteristics, there is an effect that erroneous learning can be prevented.

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

  FIG. 1 is a system configuration diagram illustrating a schematic configuration of a fuel cell system including a first embodiment of a current-voltage characteristic learning device for a fuel cell according to the present invention. This embodiment is an embodiment in which learning of the current-voltage characteristics of the fuel cell is stopped when the state of the air that is an oxidant gas as the fluid supplied to the fuel cell is less than the reference flow rate value, and is used for air flow detection. It is an Example using a flow sensor.

  In FIG. 1, a fuel cell 1 is a fuel cell in which, for example, a solid polymer electrolyte is sandwiched between an anode and a cathode as an electrolyte membrane. Hydrogen gas is supplied from a hydrogen supply means (not shown) to the anode (not shown) of the fuel cell 1 and air is supplied to the cathode by the compressor 2, and DC power is generated by the electrode reaction described below.

Anode (hydrogen electrode): H ₂ → 2H ⁺ + 2e ⁻ (1)
Cathode (oxygen electrode): 2H ⁺ + 2e ⁻ + (1/2) O ₂ → H ₂ O (2)
The air flow sensor 3 detects the flow rate of air supplied from the compressor 2 to the cathode. The power manager (PM) 4 takes out current from the fuel cell 1, performs desired power conversion, and supplies power to a load (not shown). The current sensor 5 detects the extraction current of the fuel cell 1. The voltage sensor 6 detects the total voltage of all the cells of the fuel cell 1.

  The current-voltage characteristic learning device for a fuel cell according to the present invention comprises an IV learning means 7 and a learning stop means 8.

  The IV learning unit 7 is a current-voltage characteristic learning unit that learns the current-voltage characteristic of the fuel cell 1 based on the current detected by the current sensor 5 and the voltage detected by the voltage sensor 6. Since the learning method is described as a known technique in Patent Document 1 and the like, it is omitted here.

  The learning stop unit 8 determines whether or not to stop the learning of the IV learning unit 7 according to the air flow rate reference value set according to the fuel cell target extraction current and the air flow rate detected by the air flow sensor 3. When it is determined to stop, a stop signal is sent to the IV learning means 7 to stop learning.

  Although not particularly limited, in the present embodiment, the IV learning means 7 and the learning stop means 8 are formed integrally with a controller that controls the entire fuel cell system. This controller can be composed of, for example, a microprocessor including a CPU, a ROM that stores various control parameters and tables, a working RAM, and an input / output interface.

  FIG. 2 is a flowchart for explaining the operation of the current-voltage characteristic learning apparatus according to the present embodiment, and is configured as a subroutine that is called from the main routine of the controller and executed every predetermined time. First, in step (hereinafter abbreviated as S) 201, the air flow sensor 3 detects the air flow rate. Next, in S202, it is determined whether or not the detected air flow rate is greater than or equal to a reference value. When it is determined that the value is equal to or larger than the reference value in S202, the process proceeds to S203, and current-voltage characteristic learning (IV learning) is continued. If the determination in S202 is less than the reference value, the IV characteristic may decrease due to air shortage, so the process proceeds to S204 and IV learning is stopped.

  FIG. 3 is a diagram illustrating an example of a method for setting the air flow rate reference value in the learning stop unit 8. As is clear from the cathode reaction equation (2), oxygen having a mass flow rate proportional to the extraction current of the fuel cell is consumed, so an air flow reference value proportional to the target extraction current of the fuel cell is set. Here, the proportionality constant is calculated from the excess ratio of the mass flow rate of air to the output current, the ratio of oxygen in the air, the cathode reaction equation (2), and the Faraday constant (F) determined by the characteristics of the fuel cell. Is done.

  According to the present embodiment described above, when the air flow rate detected by the flow rate sensor is less than the reference value, the learning of the current-voltage characteristic learning means is stopped. This has the effect of preventing the voltage drop from being mistakenly learned as deterioration of the fuel cell.

  Note that, in this embodiment and each embodiment described below, the same effect can be obtained by modifying the learning delay means to reduce the learning speed of the IV learning means in place of the learning stop means. Here, as a specific method of reducing the learning speed, the sampling time interval between the current value and the voltage value may be extended, or an approximate expression of the current-voltage characteristic obtained as a learning result of the current-voltage characteristic is defined. The calculation time interval of the parameters (for example, in Cited Document 1, the slope of the current-voltage characteristic and the Y intercept) may be extended.

  Next, Example 2 will be described. FIG. 4 is a system configuration diagram illustrating a schematic configuration of a fuel cell system including a second embodiment of the current-voltage characteristic learning device for a fuel cell according to the present invention. The present embodiment is an embodiment in which learning of the current-voltage characteristics of the fuel cell is stopped when the state of air, which is an oxidant gas, is less than the reference flow rate value as the fluid supplied to the fuel cell, and air is supplied. The air flow rate is estimated from the compressor speed.

  The difference between the present embodiment and the first embodiment shown in FIG. 1 is that the air flow sensor 3 that detects the air flow rate is not provided, and the air flow rate is estimated based on the rotation speed measurement value of the compressor 2. Since the other configuration is the same as that of the first embodiment shown in FIG. 1, the same components are assigned the same reference numerals and redundant description is omitted.

  The learning stop unit 8 determines whether or not to stop the learning of the IV learning unit 7 according to the compressor rotation speed reference value set according to the fuel cell target extraction current and the rotation speed measurement value of the compressor 2. .

  FIG. 5 is a flowchart for explaining the operation of the current-voltage characteristic learning device according to this embodiment. First, in step S501, the rotation speed of the compressor is detected. Next, in S502, it is determined whether or not the compressor rotation speed detection value is greater than or equal to a reference value. If it is determined in S502 that the reference value is equal to or greater than the reference value, the process proceeds to S503 and the IV learning is continued. If it is less than the reference value in the determination in S502, the process proceeds to S504, and IV learning is stopped.

  FIG. 6 is a diagram illustrating an example of a method for setting the compressor rotation speed reference value in the learning stop unit 8. Air is consumed based on the extraction current of the fuel cell, and the required compressor speed is determined. For this reason, the compressor rotation speed with respect to the air flow rate (mass flow rate) is experimentally obtained, and a compressor rotation speed reference value that can supply the air flow rate according to the target extraction current of the fuel cell is set.

  According to the present embodiment described above, since the learning of the current-voltage characteristic learning means is stopped when the air flow rate is less than the reference value, the voltage drop of the fuel cell due to the insufficient air flow rate is prevented. There is an effect that it is possible to prevent learning by mistake and deterioration.

  In this embodiment, since the air flow rate is detected by the rotation speed of the compressor, the air flow rate sensor can be omitted.

  Next, Example 3 will be described. FIG. 7 is a system configuration diagram illustrating a schematic configuration of a fuel cell system including a third embodiment of the current-voltage characteristic learning device for a fuel cell according to the present invention. The present embodiment is an embodiment in which learning of the current-voltage characteristics of the fuel cell is stopped when the state of air that is an oxidant gas as a fluid supplied to the fuel cell is less than the reference pressure value.

  The difference between the present embodiment and the first embodiment shown in FIG. 1 is that, instead of the airflow sensor 3 of the first embodiment, a pressure sensor 9 that detects the air pressure at the cathode inlet of the fuel cell 1 and the cathode of the fuel cell 1. And a pressure regulating valve 10 that regulates the cathode pressure by restricting the air at the outlet. Since the other configuration is the same as that of the first embodiment shown in FIG. 1, the same components are assigned the same reference numerals and redundant description is omitted.

  FIG. 8 is a flowchart for explaining the operation of the current-voltage characteristic learning apparatus according to this embodiment. First, in step S <b> 701, the air pressure at the cathode inlet of the fuel cell 1 is detected by the pressure sensor 9. Next, in S702, it is determined whether or not the detected air pressure at the fuel cell inlet is equal to or higher than a reference value. If the determination in S702 is greater than or equal to the reference value, the process proceeds to S703 and the IV learning is continued. If it is less than the reference value in the determination in S7023, the process proceeds to S704, and the IV learning is stopped.

  FIG. 9 is a diagram illustrating an example of a method for setting the air pressure reference value in the learning stop unit 8. An air pressure reference value is set in accordance with a preset target air pressure based on the extraction current of the fuel cell 1.

  According to the present embodiment described above, since the learning of the current-voltage characteristic learning means is stopped when the air pressure is less than the reference value, the fuel cell voltage drop caused by the shortage of air pressure is prevented. There is an effect that it is possible to prevent learning by mistake and deterioration.

  Next, Example 4 will be described. FIG. 10 is a system configuration diagram illustrating a schematic configuration of a fuel cell system including a fuel cell current-voltage characteristic learning device according to a fourth embodiment of the present invention. The present embodiment is an embodiment in which learning of the current-voltage characteristics of the fuel cell is stopped when the state of hydrogen as the fuel gas as a fluid supplied to the fuel cell is less than a reference value.

  In FIG. 10, the fuel cell 1, the power manager (PM) 4, the current sensor 5, the voltage sensor 6, the IV learning means 7, and the learning stop means 8 are a stop determination method for the learning stop means 8. Is the same as that of the first embodiment shown in FIG.

  The hydrogen tank 11 stores hydrogen as a fuel gas supplied to the fuel cell 1. The pressure sensor 12 detects the hydrogen pressure in the hydrogen tank 11. The pressure reducing valve 13 reduces the high-pressure hydrogen supplied from the hydrogen tank 11 to a predetermined medium pressure. The hydrogen pressure regulating valve 14 reduces the medium-pressure hydrogen to the operating pressure of the fuel cell 1 and supplies it to the cathode inlet of the fuel cell 1. The pressure sensor 16 detects the hydrogen pressure at the cathode inlet of the fuel cell 1. The hydrogen pressure regulating valve 14 and the pressure sensor 16 are used for cathode pressure adjustment control of the fuel cell 1.

  The hydrogen circulation pump 15 recirculates hydrogen that has not been consumed in the fuel cell 1. The purge valve 17 is a valve that is opened to manage the nitrogen concentration in the hydrogen circulation passage within a predetermined range, and discharges nitrogen accumulated in the anode and the hydrogen circulation path. The dilution fan 18 dilutes the hydrogen gas containing nitrogen discharged from the purge valve 17 to below the lower combustion limit concentration and discharges it to the outside of the system.

  The learning stop means 8 compares the hydrogen tank internal pressure detected by the pressure sensor 12, the hydrogen circulation pump rotation speed reference value set according to the fuel cell target extraction current, and the same measured value, and the fuel cell target extraction current. Whether the learning of the IV learning means 7 is to be stopped is determined according to the comparison result between the hydrogen pressure reference value set according to the measured value and the measured value.

  FIG. 11 is a flowchart for explaining the operation of the current-voltage characteristic learning device according to this embodiment. First, in S1101, the internal pressure of the hydrogen tank 11 is detected by the pressure sensor 12. Next, in S1102, it is determined whether or not the internal pressure of the hydrogen tank is equal to or higher than a reference value. If it is determined in S1102 that the value is equal to or larger than the reference value, the process proceeds to S1103, the number of rotations of the hydrogen circulation pump is detected, and the process proceeds to S1104. If it is less than the reference value in the determination in S1102, the process proceeds to S1108, and IV learning is stopped.

  In S1104, it is determined whether or not the detected value of the rotation number of the hydrogen circulation pump is equal to or greater than a reference value. If the determination in S1104 is greater than or equal to the reference value, the process proceeds to S1105, the hydrogen pressure at the fuel cell inlet is detected by the pressure sensor 16, and the process proceeds to S1106. If it is less than the reference value in the determination in S1104, the process proceeds to S1108 and IV learning is stopped.

  In S1106, it is determined whether or not the hydrogen pressure detection value at the fuel cell inlet is greater than or equal to a reference value. If it is determined in S1106 that the reference value is greater than or equal to the reference value, the process proceeds to S1107 and IV learning is continued. When it is less than the reference value in the determination in S1106, the process proceeds to S1108 and IV learning is stopped.

  FIG. 12 is a diagram illustrating an example of a method for setting the hydrogen circulation pump rotation speed reference value and the hydrogen pressure reference value in the learning stop unit 8. As is apparent from the equation (1) of the anode reaction, hydrogen having a mass flow rate proportional to the extraction current of the fuel cell is consumed. Therefore, a hydrogen flow reference value proportional to the target extraction current of the fuel cell is set. From the rotational speed characteristic of the hydrogen circulation pump 15 corresponding to this hydrogen flow rate reference value, the necessary hydrogen circulation pump rotational speed corresponding to the take-out current of the fuel cell is determined. For this reason, the hydrogen circulation pump rotational speed reference value is set according to the target extraction current of the fuel cell. Further, a hydrogen pressure reference value is set in accordance with a target hydrogen pressure that is set in advance based on the extraction current of the fuel cell.

  According to the present embodiment described above, when the hydrogen flow rate or the hydrogen pressure is less than the reference value, the learning of the current-voltage characteristic learning unit is stopped, so the voltage drop of the fuel cell due to the hydrogen shortage Can be prevented from being mistakenly learned as deterioration of the fuel cell.

  Further, according to this embodiment, since the hydrogen flow rate can be estimated from the hydrogen circulation pump rotation speed, there is an effect that the hydrogen flow rate sensor can be omitted and the configuration of the hydrogen flow path can be simplified.

  Next, Example 5 will be described. FIG. 13 is a system configuration diagram illustrating a schematic configuration of a fuel cell system including Example 5 of a current-voltage characteristic learning device for a fuel cell according to the present invention. In this embodiment, when the nitrogen concentration in the fuel circuit is out of the predetermined concentration range, it is determined that the state of hydrogen supplied to the fuel cell is less than the reference flow rate, and the current-voltage characteristics of the fuel cell are learned. This is an example of stopping.

  The difference between the present embodiment and the fourth embodiment shown in FIG. 10 is that a current sensor 19 that detects the motor drive current of the hydrogen circulation pump 15, the rotation speed measurement value of the hydrogen circulation pump 15, and the motor drive current A nitrogen concentration estimating means 20 for estimating the nitrogen concentration in the circulation path is added, and the learning stop means 8 controls the learning stop according to the nitrogen concentration estimated by the nitrogen concentration estimating means 20. The other configuration is the same as that of the fourth embodiment shown in FIG. 10, and therefore, the same components are assigned the same reference numerals and redundant description is omitted.

  The nitrogen concentration estimation means 20 estimates the nitrogen concentration from the relationship that the motor drive current increases as the nitrogen concentration increases. The learning stop unit 8 determines whether the circulation performance of the hydrogen circulation pump 15 satisfies the reference value according to the estimated nitrogen concentration value, and determines whether to stop the learning of the IV learning unit 7.

  FIG. 14 is a flowchart for explaining the operation of the current-voltage characteristic learning device according to this embodiment. First, in S1401, the number of rotations of the hydrogen circulation pump 15 is detected. Next, in S1402, the current sensor 19 detects the drive current of the hydrogen circulation pump 15. In S1403, the nitrogen concentration in the hydrogen circulation system is estimated from the rotation speed of the hydrogen circulation pump 15 and the motor drive current. Next, in S1404, it is determined whether or not the estimated nitrogen concentration value in the hydrogen circulation system is within the reference range between the reference upper limit value and the reference lower limit value. If it is determined in S1404 that it is within the reference range, the process proceeds to S1405 and the IV learning is continued. If it is determined in S1404 that it is out of the reference range, the process proceeds to S1406 and IV learning is stopped.

  FIG. 15 is a diagram illustrating an example of a method for setting the nitrogen concentration reference value in the hydrogen circulation system in the learning stop unit 8. The nitrogen concentration range is set so that the circulation performance of the hydrogen circulation pump satisfies the required circulation flow rate. Nitrogen molecules are about 14 times heavier than hydrogen molecules, so if the nitrogen concentration is high, the gas in the circulation system becomes heavier and the circulation performance of the hydrogen circulation pump 15 is reduced. Further, even if the nitrogen concentration is too low, the gas in the circulation system becomes light, and a phenomenon that the hydrogen circulation pump 15 rotates idly occurs and the circulation performance deteriorates. Accordingly, if the nitrogen concentration is taken on the horizontal axis and the circulation performance of the hydrogen circulation pump is taken on the vertical axis, the characteristic diagram has an upwardly convex shape.

  According to the present embodiment described above, when the nitrogen concentration or hydrogen concentration of the hydrogen circulation system is out of the concentration range in which the hydrogen circulation device performs the circulation function, it is determined that the hydrogen flow rate does not satisfy the reference value, Since the learning of the current-voltage characteristic learning means is stopped, there is an effect that it is possible to prevent the fuel cell voltage drop caused by the insufficient hydrogen flow rate from being mistakenly learned as the deterioration of the fuel cell.

  Next, Example 6 will be described. FIG. 16 is a system configuration diagram illustrating a schematic configuration of a fuel cell system including a sixth embodiment of the current-voltage characteristic learning device for a fuel cell according to the present invention. The present embodiment is an embodiment in which learning of the current-voltage characteristics of the fuel cell is stopped when the state of hydrogen as the fuel gas as the fluid supplied to the fuel cell is less than the reference pressure value.

  The difference between the present embodiment and the fourth embodiment shown in FIG. 10 is that hydrogen is circulated by an ejector 21 which is a fluid pump instead of the hydrogen circulation pump 15 as a hydrogen circulation means. The other configuration is the same as that of the fourth embodiment shown in FIG. 10, and therefore, the same components are assigned the same reference numerals and redundant description is omitted.

  The learning stop means 8 of this embodiment determines whether or not to stop learning of the IV learning means 7 depending on whether the hydrogen pressure at the fuel cell inlet satisfies the reference value.

  FIG. 17 is a flowchart for explaining the operation of the current-voltage characteristic learning device according to this embodiment. First, in S1701, the pressure sensor 16 detects the hydrogen pressure at the fuel cell inlet. Next, in S1702, it is determined whether or not the hydrogen pressure detection value at the fuel cell inlet is greater than or equal to a reference value. If it is determined in S1702 that the value is equal to or greater than the reference value, the process proceeds to S1703, and the IV learning is continued. If it is less than the reference value in the determination in S1702, the process proceeds to S1704 and IV learning is stopped.

  FIG. 18 is a diagram illustrating an example of a method for setting the hydrogen pressure reference value in the learning stop unit 8. A hydrogen pressure reference value is set in accordance with a target hydrogen pressure set in advance based on the extraction current of the fuel cell. When the hydrogen pressure is less than the reference value, the hydrogen supply amount is insufficient or the circulation performance of the ejector may be deteriorated, so that the IV characteristic may be lowered. Stop learning.

  According to the present embodiment described above, when the hydrogen pressure is less than the reference value, it is determined that the hydrogen flow rate is less than the reference value, and learning of the current-voltage characteristic learning means is stopped. In addition, there is an effect that it is possible to prevent the fuel cell voltage drop caused by the insufficient hydrogen flow rate from being mistakenly learned as the deterioration of the fuel cell.

  Next, Example 7 will be described. FIG. 19 is a system configuration diagram illustrating a schematic configuration of a fuel cell system including a seventh embodiment of the current-voltage characteristic learning device for a fuel cell according to the present invention. The present embodiment is an embodiment in which learning of the current-voltage characteristics of the fuel cell is stopped when the humidified state of the air, which is the oxidant gas, is less than the reference value as the fluid supplied to the fuel cell, and the humidified state of the air Is determined based on the air temperature at the cathode inlet and outlet.

  The difference between the present embodiment and the third embodiment shown in FIG. 7 is that a temperature sensor 22 that detects the air temperature at the cathode inlet of the fuel cell 1 and a temperature sensor 23 that detects the air temperature at the cathode outlet of the fuel cell 1. The pump 24 for circulating the coolant, the radiator 25 for releasing the heat of the coolant to the outside of the system, the fan 26 that blows air to the radiator 25 and the air compressed by the compressor 2 are cooled with the coolant, and the cathode of the fuel cell 1 The air cooler 34 to be supplied to is added. Since the other configuration is the same as that of the third embodiment shown in FIG. 7, the same components are assigned the same reference numerals and redundant description is omitted.

  The temperature sensors 22 and 23 detect the air temperatures at the cathode inlet and the cathode outlet of the fuel cell 1, respectively. The learning stop unit 8 determines the humidification state of the fuel cell based on whether or not each of the cathode inlet and outlet air temperatures of the fuel cell 1 is below the reference value, and based on this determination, the IV learning unit 7 Controls whether to stop learning.

  FIG. 20 is a flowchart for explaining the operation of the current-voltage characteristic learning device according to this embodiment. First, in S2001, the temperature sensor 22 detects the air temperature at the cathode inlet. Next, in S2002, it is determined whether or not the cathode inlet air temperature detection value is below a reference value. When it is below the reference value in the determination in S2002, the process proceeds to S2003, the temperature temperature of the cathode outlet is detected by the temperature sensor 23, and the process proceeds to S2004. If the reference value is exceeded in the determination in S2002, it is determined that the air supplied to the cathode is insufficiently humid due to malfunction of the air cooler 34, etc., and the process proceeds to S2006 to stop the IV learning.

  In S2004, it is determined whether or not the air temperature detection value at the cathode outlet is below a reference value. When it is below the reference value in the determination in S2004, the process proceeds to S2005 and IV learning is continued. If the reference value is exceeded in the determination of S2004, the temperature at the cathode outlet rises due to the cooling capacity of the fuel cell 1 and the amount of moisture taken out by wave sword exhaust increases, and it is determined that the humidification is insufficient. Advance IV learning is stopped. The reference value for determination in S2002 and the reference value for determination in S2004 are set by experimentally determining a temperature threshold value that does not dry the electrolyte membrane of the fuel cell 1.

  According to the present embodiment described above, since the learning of the current-voltage characteristic learning means is stopped when the humidification state of the fuel cell is less than the reference value, the fuel cell caused by insufficient humidification of the electrolyte membrane Thus, it is possible to prevent the voltage drop from being mistakenly learned as deterioration of the fuel cell.

  Further, according to the present embodiment, since the humidified state is estimated from the air temperature at the cathode inlet or outlet, the humidified state can be estimated using a temperature sensor that is easy to handle and low cost instead of the humidity sensor. There is an effect.

  Next, Example 8 will be described. FIG. 21 is a system configuration diagram illustrating a schematic configuration of a fuel cell system including an eighth embodiment of the current-voltage characteristic learning device for a fuel cell according to the present invention. The present embodiment is an embodiment in which learning of the current-voltage characteristics of the fuel cell is stopped when the humidified state of the air, which is the oxidant gas, is less than the reference value as the fluid supplied to the fuel cell, and the humidified state of the air Is determined based on the number of rotations of the pump 24 for circulating the coolant, the coolant pressure at the inlet of the fuel cell 1, and the coolant temperature at the outlet of the fuel cell 1.

  The difference between the present embodiment and the seventh embodiment shown in FIG. 19 is that a pressure sensor 28 that detects the coolant pressure at the inlet of the fuel cell 1 and a temperature sensor 27 that detects the coolant temperature at the outlet of the fuel cell 1 Is added to the learning stop means 8 together with the detection values of these sensors together with the measured value of the rotational speed of the pump 24 for circulating the coolant. Since other configurations are the same as those of the seventh embodiment shown in FIG. 19, the same components are assigned the same reference numerals, and redundant descriptions are omitted.

  Whether the learning stop means 8 stops the learning of the IV learning means 7 depending on whether each of the coolant temperature at the fuel cell outlet, the coolant pump rotational speed, and the fuel cell inlet coolant pressure satisfies the reference values. Judge whether or not.

  FIG. 22 is a flowchart for explaining the operation of the current-voltage characteristic learning device according to this embodiment. First, in S2201, the temperature sensor 27 detects the coolant temperature at the outlet of the fuel cell. Next, in S2202, it is determined whether the fuel cell outlet coolant temperature is below a reference value. When it is below the reference value in the determination in S2202, the process proceeds to S2203, the number of revolutions of the coolant pump 24 is detected, and the process proceeds to S2204. If the reference value is exceeded in the determination in S2202, it is determined that the temperature of the fuel cell 1 is too high and tends to dry, the process proceeds to S2208, and IV learning is stopped.

  In S2204, it is determined whether or not the rotation speed detection value of the coolant pump 24 is greater than or equal to a reference value. If the determination in S2204 is equal to or greater than the reference value, the process proceeds to S2205, the coolant pressure at the fuel cell inlet is detected by the pressure sensor 28, and the process proceeds to S2206. If it is less than the reference value in the determination in S2204, it is determined that the coolant circulation amount is insufficient, and the process proceeds to S2208 and IV learning is stopped.

  In S2206, it is determined whether or not the detected coolant pressure value at the fuel cell inlet is greater than or equal to a reference value. If it is determined in step S2206 that the reference value is equal to or greater than the reference value, the process proceeds to step S2207 and IV learning is continued. If it is less than the reference value in the determination in S2206, it is determined that the coolant circulation amount is insufficient, and the process proceeds to S2208 and IV learning is stopped.

  FIG. 23 is a diagram illustrating an example of a method for setting the coolant pump rotation speed reference value and the coolant pressure reference value in the learning stop unit 8. For example, when the coolant pump rotational speed is set based on the target extraction current of the fuel cell, the reference value is also set according to the target extraction current of the fuel cell. Since the coolant pump rotation speed is determined and the coolant pressure is determined based on the target extraction current of the fuel cell, the coolant pressure reference value is set according to the target extraction current.

  According to the present embodiment described above, when the coolant pump rotational speed is less than the reference value, the coolant pressure is less than the reference value, or the coolant temperature exceeds the reference value, the current-voltage Since the learning of the characteristic learning means is stopped, the fact that the fuel cell voltage has dropped due to the deterioration of the humidification state of the fuel cell due to insufficient circulation of the coolant is mistakenly learned as the deterioration of the fuel cell. There is an effect that it can be prevented.

  Next, Example 9 will be described. FIG. 24 is a system configuration diagram illustrating a schematic configuration of a fuel cell system including a ninth embodiment of the current-voltage characteristic learning device for a fuel cell according to the present invention. This embodiment is an embodiment in which learning of the current-voltage characteristics of the fuel cell is stopped when the state of pure water for humidification as the fluid supplied to the fuel cell is less than the reference value.

  In FIG. 24, the fuel cell 1, the power manager (PM) 4, the current sensor 5, the voltage sensor 6, the IV learning means 7, and the learning stop means 8 are a stop determination method for the learning stop means 8. Is the same as that of the first embodiment shown in FIG.

  The pump 29 is a pump that supplies the fuel cell 1 with pure water for humidifying the fuel cell. The pure water tank 30 stores the pure water for humidification, supplies the pure water to the pump 29, and accepts excess pure water discharged from the fuel cell 1. The water level sensor 31 detects the water level of the pure water tank 30. The pressure sensor 32 detects the pure water pressure at the fuel cell inlet.

  Whether the learning stop means 8 stops the learning of the IV learning means 7 depending on whether the water level of the pure water tank, the rotational speed of the pure water pump 29, and the pure water pressure at the fuel cell inlet satisfy the reference values. Judge whether or not.

  FIG. 25 is a flowchart for explaining the operation of the current-voltage characteristic learning device according to this embodiment. First, in S2501, the water level sensor 31 detects the water level of the pure water tank. Next, in S2502, it is determined whether or not the water level in the pure water tank is equal to or higher than a reference value. If the determination in S2502 is greater than or equal to the reference value, the process proceeds to S2503 to detect the rotational speed of the pure water pump 29. If the determination in S2202 is less than the reference value, the water balance is insufficient and the fuel cell 1 is in a dry state from the reference, and the process proceeds to S2508 and IV learning is stopped.

  In S2504, it is determined whether the rotation speed detection value of the pure water pump 29 is equal to or greater than a reference value. If the determination in S2504 is greater than or equal to the reference value, the process proceeds to S2505, the pressure sensor 32 detects the pure water pressure at the fuel cell inlet, and the process proceeds to S2506. If it is less than the reference value in the determination in S2504, the pure water pump 29 is insufficiently rotated, and the process proceeds to S2508 and IV learning is stopped.

  In S2506, it is determined whether or not the pure water pressure detection value at the fuel cell inlet is greater than or equal to a reference value. If it is determined in step S2506 that the reference value is equal to or greater than the reference value, the process proceeds to step S2507 to continue IV learning. If it is less than the reference value in the determination in S2506, it is determined that the pure water pressure is insufficient, and the process proceeds to S2508 and IV learning is stopped.

  FIG. 26 is a diagram illustrating an example of a method for setting the pure water pump rotation speed reference value and the pure water pressure reference value in the learning stop unit 8. For example, when the rotation speed of the pure water pump is set based on the target extraction current of the fuel cell, the reference value is also set according to the target extraction current of the fuel cell. Since the pure water pump speed is determined and the pure water pressure is determined based on the target extraction current of the fuel cell, the pure water pressure reference value is set according to the target extraction current.

  According to the present embodiment described above, when the pure water level is less than the reference value, or when the pure water pump rotational speed is less than the reference value, the learning of the current-voltage characteristic learning means is stopped. There is an effect that it is possible to prevent the fact that the voltage of the fuel cell resulting from the insufficient humidification of the fuel cell is mistakenly learned as the deterioration of the fuel cell.

  Next, Example 10 will be described. FIG. 27 is a system configuration diagram illustrating a schematic configuration of a fuel cell system including a fuel cell current-voltage characteristic learning device according to a tenth embodiment of the present invention. This embodiment stops learning of the current-voltage characteristics of the fuel cell when either the cell voltage or the cell column voltage of the fuel cell is lower than the average cell voltage or the average cell column voltage by a reference value or more. It is.

  In FIG. 24, the fuel cell 1, the power manager (PM) 4, the current sensor 5, the voltage sensor 6, the IV learning means 7, and the learning stop means 8 are a stop determination method for the learning stop means 8. Is the same as that of the first embodiment shown in FIG.

  The cell string voltage sensor 32 detects the voltage for each cell string in which a plurality of cells are connected in series over the entire fuel cell 1. When the fuel cell 1 includes a plurality of substacks, the cell train may be the same as the substack, and the cell train voltage sensor may be a sensor that detects the total voltage of each substack. The cell voltage sensor 33 detects the voltage of each cell over the entire fuel cell 1.

  The learning stop unit 8 determines whether or not to stop the learning of the IV learning unit 7 according to the cell string voltage detection value and the cell voltage detection value.

  FIG. 28 is a flowchart for explaining the operation of the current-voltage characteristic learning device according to this embodiment. First, in S2801, the cell column voltage sensor 32 detects the voltage for each cell column. In step S2802, it is determined whether any cell column voltage is equal to or less than (average cell column voltage−reference value). If any cell column voltage is less than (average cell column voltage-reference value) in the determination of S2802, it is determined that an abnormality has occurred in the gas supply system to the cell column and the cell column voltage has decreased. In step S2806, the IV learning is stopped. If none of the cell column voltages are below (average cell column voltage-reference value), that is, if all the cell column voltages exceed (average cell column voltage-reference value), the cell voltage is detected in S2803. .

  Following S2803, it is determined in S2804 whether any cell voltage is equal to or less than (average cell voltage-reference value). If any cell voltage is less than (average cell voltage-reference value), it is determined that the cell is clogged with water and the cell voltage is lowered, and IV learning is stopped in S2806. . If any cell voltage is not less than (average cell column voltage-reference value), that is, if all cell voltages exceed (average cell column voltage-reference value), IV learning is continued in S2805. To do.

  Here, for each reference value, a voltage value that can detect a cell row voltage drop due to an abnormality in a gas supply system such as a manifold or a cell voltage drop due to water clogging is experimentally obtained and set.

  In this embodiment, in S2804, it is determined whether or not any cell voltage is equal to or less than (average cell voltage−reference value). Instead, a specific cell voltage (average cell voltage− You may change so that it may determine whether it is below a reference value). In this case, the specific cell voltage to be determined is the voltage of a cell that is likely to be clogged, such as a cell having the lowest temperature on the cooling water inlet side at the end of the fuel cell stack or a cell on the most downstream side of the gas supply path. And

  According to the present embodiment described above, when any cell column voltage decreases or when any cell voltage decreases, the learning of the current-voltage characteristic learning means is stopped. There is an effect that it is possible to prevent erroneous learning of a voltage drop of the fuel cell due to a gas supply system or water clogging in the cell as a voltage drop due to deterioration of the fuel cell.

1 is a system configuration diagram showing Embodiment 1 of a current-voltage characteristic learning device for a fuel cell according to the present invention. 3 is a flowchart for explaining the operation of the first embodiment. It is a figure explaining the setting method of the air flow rate reference value in Example 1. FIG. It is a system block diagram which shows Example 2 of the current-voltage characteristic learning apparatus of the fuel cell which concerns on this invention. 10 is a flowchart for explaining the operation of the second embodiment. It is a figure explaining the setting method of the compressor rotation speed reference value in Example 2. FIG. It is a system block diagram which shows Example 3 of the current-voltage characteristic learning apparatus of the fuel cell which concerns on this invention. 10 is a flowchart for explaining the operation of the third embodiment. 6 is a diagram illustrating a method for setting a fuel cell inlet air pressure reference value in Embodiment 3. FIG. It is a system block diagram which shows Example 4 of the current-voltage characteristic learning apparatus of the fuel cell which concerns on this invention. 10 is a flowchart for explaining the operation of the fourth embodiment. It is a figure explaining the setting method of the hydrogen circulation pump rotation speed reference value and fuel cell inlet hydrogen pressure reference value in Example 4. FIG. It is a system block diagram which shows Example 5 of the current-voltage characteristic learning apparatus of the fuel cell which concerns on this invention. 10 is a flowchart for explaining the operation of the fifth embodiment. It is a figure explaining the setting method of the circulation performance reference | standard range of the hydrogen circulation pump in Example 5. FIG. It is a system block diagram which shows Example 6 of the current-voltage characteristic learning apparatus of the fuel cell which concerns on this invention. 10 is a flowchart for explaining the operation of the sixth embodiment. It is a figure explaining the setting method of the fuel cell inlet hydrogen pressure reference value in Example 6. FIG. It is a system block diagram which shows Example 7 of the current-voltage characteristic learning apparatus of the fuel cell which concerns on this invention. 10 is a flowchart for explaining the operation of the seventh embodiment. It is a system block diagram which shows Example 8 of the current-voltage characteristic learning apparatus of the fuel cell which concerns on this invention. 10 is a flowchart illustrating the operation of an eighth embodiment. It is a figure explaining the setting method of the coolant pump rotation speed reference value and fuel cell inlet coolant pressure reference value in Example 8. FIG. It is a system block diagram which shows Example 9 of the current-voltage characteristic learning apparatus of the fuel cell which concerns on this invention. 22 is a flowchart for explaining the operation of the ninth embodiment. It is a figure explaining the setting method of the pure water pump rotation speed reference value and fuel cell inlet pure water pressure reference value in Example 9. FIG. It is a system block diagram which shows Example 10 of the current-voltage characteristic learning apparatus of the fuel cell which concerns on this invention. 18 is a flowchart for explaining the operation of the tenth embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Fuel cell 2 ... Compressor 3 ... Air flow sensor 4 ... Power manager 5 ... Current sensor 6 ... Voltage sensor 7 ... IV learning means 8 ... Learning stop means 9 ... Pressure sensor 10 ... Pressure regulating valve 11 ... Hydrogen tank 12 ... Pressure Sensor 13 ... Pressure reducing valve 14 ... Hydrogen pressure regulating valve 15 ... Hydrogen circulation pump 16 ... Pressure sensor 17 ... Purge valve 18 ... Dilution fan 19 ... Current sensor 20 ... Nitrogen concentration estimating means
21 ... Ejector 22 ... Temperature sensor 23 ... Temperature sensor 24 ... Pump 25 ... Radiator 26 ... Fan 27 ... Temperature sensor 28 ... Pressure sensor 29 ... Pump 30 ... Pure water tank 31 ... Water level sensor 32 ... Cell string voltage sensor 33 ... Cell voltage Sensor 34 ... Air cooler

Claims (19)

In a current-voltage characteristic learning device of a fuel cell that learns a current-voltage characteristic of a fuel cell that generates electricity by an electrochemical reaction between a fuel gas and an oxidant gas through an electrolyte membrane sandwiched between an anode and a cathode,
Current-voltage characteristic learning means for learning the current-voltage characteristic of the fuel cell by inputting the current value and voltage value of the fuel cell;
A learning stop means for stopping the learning of the current-voltage characteristic learning means when the state of any one of the fuel gas, oxidant gas, coolant, and humidifying water supplied to the fuel cell deviates from a desired operating state; ,
An apparatus for learning current-voltage characteristics of a fuel cell, comprising: The fluid is air as an oxidant gas,
An air flow rate detecting means for detecting or estimating an air flow rate supplied to the fuel cell;
2. The learning stop unit stops learning of the current-voltage characteristic learning unit when the air flow rate value detected or estimated by the air flow rate detection unit is less than a reference flow rate value. Current-voltage characteristic learning device for fuel cell.   3. The fuel cell current-voltage characteristic learning device according to claim 2, wherein the air flow rate detecting means detects an air flow rate by a flow rate sensor.   3. The fuel cell current-voltage characteristic learning device according to claim 2, wherein the air flow rate detecting means estimates an air flow rate from a rotation speed of a compressor that supplies air to the fuel cell. The fluid is air as an oxidant gas,
Air pressure detecting means for detecting the air pressure supplied to the fuel cell;
2. The fuel according to claim 1, wherein the learning stop unit stops learning of the current-voltage characteristic learning unit when the air pressure value detected by the air pressure detection unit is less than a reference pressure value. A battery current-voltage characteristic learning device. The fluid is hydrogen as a fuel gas;
A hydrogen flow rate detection means for detecting or estimating the flow rate of hydrogen supplied to the fuel cell;
2. The learning stop unit stops learning of the current-voltage characteristic learning unit when the hydrogen flow rate value detected or estimated by the hydrogen flow rate detection unit is less than a reference flow rate value. Current-voltage characteristic learning device for fuel cell. A hydrogen circulation path and a hydrogen circulation pump for circulating unreacted hydrogen discharged from the anode to the anode;
A hydrogen circulation pump rotation speed detection means for detecting the rotation speed of the hydrogen circulation pump;
7. The current-voltage characteristic learning device for a fuel cell according to claim 6, wherein the hydrogen flow rate detecting means estimates a hydrogen flow rate based on the rotational speed detected by the hydrogen circulation pump rotational speed detecting means. Nitrogen concentration detection means for detecting or estimating the nitrogen concentration or hydrogen concentration of the hydrogen circuit,
The hydrogen flow rate detection means is configured such that when the detected or estimated hydrogen concentration or hydrogen concentration of the hydrogen circulation path is out of a concentration range in which the hydrogen circulation device performs a circulation function, the hydrogen flow rate value is less than a reference flow rate value. The current-voltage characteristic learning device for a fuel cell according to claim 7, wherein: Comprising hydrogen pressure detecting means for detecting the pressure of hydrogen supplied by the anode;
7. The hydrogen flow rate detection unit determines that the hydrogen flow rate value is less than a reference flow rate value when a hydrogen pressure value detected by the hydrogen pressure detection unit is less than a reference pressure value. The fuel cell current-voltage characteristic learning device according to claim. Hydrogen tank pressure detection means for detecting the internal pressure of the hydrogen tank is provided,
The hydrogen flow rate detection unit determines that the hydrogen flow rate value is less than a reference flow rate value when the hydrogen tank internal pressure detected by the hydrogen tank pressure detection unit is less than a pressure at which hydrogen supply is possible. The current-voltage characteristic learning apparatus for a fuel cell according to claim 6. A humidifying state detecting means for detecting or estimating the humidifying state of the gas supplied to the fuel cell;
2. The fuel cell current according to claim 1, wherein the learning stopping unit stops learning of the current-voltage characteristic learning unit when the detected or estimated humidified state is less than a reference value. 3. Voltage characteristic learning device. Air temperature detection means for detecting the air temperature at the cathode inlet or outlet,
12. The fuel cell current-voltage characteristic learning device according to claim 11, wherein the humidified state detecting means estimates the humidified state from the detected cathode inlet or outlet air temperature. A coolant pump rotational speed detection means for detecting the rotational speed of the coolant pump;
12. The fuel cell current-voltage characteristic learning device according to claim 11, wherein when the coolant pump rotation speed is less than a reference value, the humidification state detection means determines that the humidification state is less than the reference value. . A coolant pressure detecting means for detecting the pressure of the coolant is provided,
12. The current-voltage characteristic of a fuel cell according to claim 11, wherein when the detected coolant pressure is less than a reference value, the humidification state detection means determines that the humidification state is less than a reference value. Learning device. A coolant temperature detecting means for detecting the temperature of the coolant is provided,
12. The current-voltage characteristic of a fuel cell according to claim 11, wherein when the detected coolant temperature exceeds a reference value, the humidification state detection means determines that the humidification state does not satisfy the reference value. Learning device. The fluid is pure water for humidification supplied to the fuel cell,
The fuel cell system includes a pure water tank that stores pure water supplied to the fuel cell, a water level sensor that detects a water level of the pure water tank, and a pure water pump that supplies pure water from the pure water tank to the fuel cell. ,
The learning stop means stops learning of the current-voltage characteristic learning means when the detection value of the water level sensor is less than a predetermined value or when the pure water circulation amount by the pure water pump is less than a reference flow rate value. The current-voltage characteristic learning device for a fuel cell according to claim 1. Cell voltage detection means for detecting the voltage of each cell of the fuel cell over the entire fuel cell;
2. The learning stop unit stops learning of the current-voltage characteristic learning unit when any one of the detected cell voltages is reduced by a reference value or more. Current-voltage characteristic learning device for fuel cell. A cell string voltage detecting means for detecting a voltage for each cell string in which a plurality of cells of the fuel cell are connected in series over the entire fuel cell;
2. The learning stop unit stops learning of the current-voltage characteristic learning unit when any one of the detected cell column voltages decreases by a reference value or more. The current-voltage characteristic learning device for a fuel cell according to claim 1. In a current-voltage characteristic learning device of a fuel cell that learns a current-voltage characteristic of a fuel cell that generates electricity by an electrochemical reaction between a fuel gas and an oxidant gas through an electrolyte membrane sandwiched between an anode and a cathode,
Current-voltage characteristic learning means for learning the current-voltage characteristic of the fuel cell by inputting the current value and voltage value of the fuel cell;
Learning delay means for reducing the learning speed of the current-voltage characteristic learning means when the state of any one of the fuel gas, oxidant gas, coolant, and humidifying water supplied to the fuel cell deviates from a desired operating state. When,
An apparatus for learning current-voltage characteristics of a fuel cell, comprising:

Images