JP6223311B2 - Control method of fuel cell system - Google Patents

Description

  The present invention relates to a control method for a fuel cell system. More specifically, the present invention relates to a control method for a fuel cell system that appropriately adjusts the state of moisture inside the fuel cell during power generation.

  The fuel cell stack is configured by stacking a plurality of fuel cells. Each fuel cell is composed of a membrane electrode structure (MEA) sandwiched between a pair of separators. The membrane electrode structure is composed of two electrodes, an anode electrode and a cathode electrode, and a solid polymer electrolyte membrane taught in these electrodes. In this fuel cell stack, when hydrogen gas is supplied to the anode flow path formed on the anode electrode side and air is supplied to the cathode flow path formed on the cathode electrode side, power is generated and water is generated by these reactions. Generated. The generated water wets the electrolyte membrane and flows out to the gas flow path through which the air, hydrogen gas, and the like flow.

  Patent Document 1 describes an invention in which an internal moisture state of a fuel cell is maintained in an appropriate state using an impedance meter. In the invention of Patent Document 1, using the fact that there is a correlation between the wet state of the electrode sandwiching the electrolyte membrane and the impedance resistance value of the fuel cell, the impedance meter detects the state where the electrode is too wet. Increases the flow rate of the gas supplied to the electrode.

JP 7-235324 A

  In the invention of Patent Document 1, after increasing the flow rate of the gas supplied to the electrode as described above, the impedance meter detects that the electrode is too wet (ie, the impedance resistance value is a predetermined value). In response to returning to the normal range), the gas flow rate is restored to the normal amount.

  However, when the flow rate of the gas is restored to the normal amount at the timing as in the invention of Patent Document 1, the water staying in the gas flow path inside the fuel cell (hereinafter, such water is referred to as “retaining water”). In some cases, so-called flooding may occur. This means that even if the impedance resistance value is maintained within the normal range, the amount of accumulated water in the gas flow path within the fuel cell is not necessarily maintained within an appropriate range.

  An object of the present invention is to provide a control method of a fuel cell system that can maintain the moisture state inside the fuel cell in an appropriate state.

(1) A fuel cell system (for example, a fuel cell system 1 described later) includes a fuel cell (for example, a stack 2 described later) that generates power when an oxidant gas and a fuel gas are supplied, and an oxidant gas to the fuel cell. An oxidant gas supply device (for example, a cathode system 3 described later), a fuel gas supply device (for example, an anode system 4 described later) for supplying fuel gas to the fuel cell, and an impedance of the fuel cell are detected An impedance detection device (for example, an impedance sensor 24 described later). The power generation of the fuel cell is performed under a normal power generation mode in which power generation control is performed under a predetermined reference power generation condition, and under a dry power generation condition determined so that the membrane of the fuel cell is dried more than the reference power generation condition. Dry power generation mode for performing power generation control, and stagnant water discharge for performing power generation control under discharge power generation conditions determined so that more stagnant water in the gas flow path of the fuel cell is discharged than the reference power generation conditions Three power generation modes are defined: a power generation mode. The fuel cell system control method according to the present invention includes an over-humidity determining step of determining whether or not the fuel cell is in an over-humidified state based on a detection value of the impedance detection device (for example, S31 in FIG. 7, FIG. 9 S53), and a dry power generation process for generating power in the dry power generation mode until it is determined not to be in an excessively humid state after being determined as being in an excessively humid state in the excessive humidification determining process (for example, FIG. 34, FIG. 8 from t2 to t4 and t6 to t8), a request acquisition step for acquiring a required value for the output of the fuel cell (for example, S15 in FIG. 4), and detection of the impedance detection device after the dry power generation step is completed. value predetermined lower threshold (e.g., retained water determination threshold described later) latency state in which exceeds the stipulated in accordance with the required value described above acquired (e.g., t8 to t9 in FIG. 8) or is maintained Retained water discharge step for generating power by the accumulated water discharge power mode (e.g., S36 of FIG. 7, t4-t6 and t8~t9 in FIG. 8), characterized in that it comprises a a.

(2) In this case, in the previous SL accumulated water discharge step, it is preferable to lengthen the waiting time as the required value the acquired decreases.

(3) In this case, the control method includes an output decrease rate determination step (for example, S22 in FIG. 6) for determining whether or not the decrease rate of the output of the fuel cell is greater than a predetermined output decrease rate threshold, If the decrease rate is larger than the output decrease rate threshold value, a drainage priority step (for example, S52 in FIG. 9) that performs power generation in the accumulated water power generation / discharge mode regardless of the determination result of the overhumidification determination step It is preferable to provide.

(4) In this case, the control method includes a voltage drop determination step (for example, S23 in FIG. 6) for determining whether or not the voltage of the fuel cell is equal to or lower than a predetermined voltage threshold, and the voltage is the voltage threshold. In the case of the following, it is preferable to further include a drainage priority step (for example, S52 in FIG. 9) that generates power in the accumulated water discharge power generation mode regardless of the determination result of the overhumidification determination step.

  (5) In this case, the control method further includes a warm-up determination step (for example, S33 in FIG. 7 and S55 in FIG. 9 described later) for determining whether or not the fuel cell has been warmed up. When the warm-up is not completed, it is preferable to perform the stagnant water discharge step regardless of the determination result of the excessive humidification determination step.

(1) In the present invention, when it is determined that the state is an excessively humidified state based on the detection value of the impedance detection device, power generation is performed in the dry power generation mode until it is determined that the state is not the excessively humidified state thereafter. Thereby, the water content of the membrane of the fuel cell, which is considered to be mainly correlated with the detection value of the impedance detection device, can be maintained at an appropriate amount. By the way, when the water present in the fuel cell is divided into those contained in the membrane (containing the membrane water) and those contained in the gas flow path (retaining water), the remaining water exceeded the water retention limit of the membrane. It may be considered that the water content of the membrane gradually oozes into the flow channel over time. In the present invention, in consideration of such a generation model of stagnant water, a state in which the detection value of the impedance detection device exceeds a predetermined lower limit threshold even after it is determined not to be in an excessively humid state as described above is a predetermined waiting time. Power is generated in the accumulated water discharge power generation mode until the time is maintained. Thus, by performing power generation in the stagnant water discharge power generation mode over the waiting time, the stagnant water generated by oozing from the membrane can be reliably discharged. As described above, according to the present invention, it is possible to maintain both the amount of membrane water contained in the fuel cell and the amount of accumulated water during power generation at appropriate amounts.

  (2) Since the amount of fuel gas and oxidant gas supplied to the fuel cell tends to decrease as the required value for the output of the fuel cell decreases, the accumulated water is discharged from the flow path inside the fuel cell. There is a tendency for the time taken to become longer. In the present invention, the retained water can be reliably discharged by increasing the waiting time as the required value becomes smaller.

(3) As described above, in the present invention, on the premise that the retained water is generated by oozing out the membrane water content, basically, after performing the dry power generation mode with emphasis on the drying of the membrane, The stagnant water discharge power generation mode will be implemented with emphasis on the stagnant water discharge. On the other hand, if the output of the fuel cell rapidly decreases, the amount of accumulated water in the fuel cell may also increase abruptly. In such a case, the accumulated water is discharged as quickly as possible to prevent flooding. It is preferable. Therefore, in the present invention, when the rate of decrease in the output of the fuel cell is larger than a predetermined threshold, the determination result of the overhumidification determination step is used to reverse the execution priority of the dry power generation mode and the stagnant water discharge power generation mode. Regardless of power generation in the stagnant water discharge power generation mode. Thereby, flooding can be prevented in advance.

(4) If the accumulated water inside the fuel cell during power generation increases excessively, the voltage may drop significantly (so-called flooding). In the present invention, when the voltage of the fuel cell becomes equal to or lower than a predetermined threshold value, the over-humidity determination is performed so as to reverse the execution priority of the dry power generation mode and the stagnant water discharge power generation mode as in the case of the invention of (3). Power generation is performed in the accumulated water discharge power generation mode regardless of the determination result of the process. Thereby, stagnant water can be discharged quickly and the power generation of the fuel cell can be stabilized.

  (5) While performing the warm-up control, it has not been so long since the start of the fuel cell system that had been in a low temperature environment until then, and the amount of moisture inside the fuel cell is appropriate. Since it is slightly larger than the amount, flooding is likely to occur. In the present invention, in consideration of this point, when the warm-up is not completed, the stagnant water discharge step is performed regardless of the determination result of the overhumidification determination step. Thereby, stagnant water can be discharged quickly, flooding can be prevented, and the power generation of the fuel cell at the start can be stabilized.

It is a figure which shows the structure of the fuel cell system which concerns on one Embodiment of this invention. It is a flowchart which shows the procedure of the warming-up control at the time of starting of a fuel cell system. It is a figure for demonstrating the content of the production | generation and discharge | emission model of the stagnant water of a stack. It is a flowchart which shows the procedure of the process which determines the presence or absence of stagnant water. It is the table | surface which compared four types of electric power generation conditions from the functional surface. It is a flowchart which shows the procedure of an electric power generation condition selection process. It is a flowchart which shows the procedure of the electric power generation condition selection process at the time of water content control priority. It is a time chart which shows an example at the time of selecting electric power generation conditions under the process of FIG. It is a flowchart which shows the procedure of the electric power generation condition selection process at the time of stagnant water control priority.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a diagram showing a configuration of a fuel cell system 1 according to the present embodiment.
The fuel cell system 1 includes a fuel cell stack 2, an anode system 3 that supplies hydrogen as a reaction gas to the fuel cell stack 2, and a cathode system 4 that supplies air containing oxygen as a reaction gas to the fuel cell stack 2. And a cooling device 5 that cools the fuel cell stack 2, a battery B that stores electric power generated by the fuel cell stack 2, and a tire (not shown) is driven by the supply of electric power from the fuel cell stack 2 and the battery B A travel motor M and an ECU 6 as these electronic control units are provided. The fuel cell system 1 is mounted on a fuel cell vehicle (not shown) using the tire as a driving wheel.

  The fuel cell stack (hereinafter simply referred to as “stack”) 2 has a stack structure in which, for example, several tens to several hundreds of cells are stacked. Each fuel cell is configured by sandwiching a membrane electrode structure (MEA) between a pair of separators. The membrane electrode structure is composed of two electrodes, an anode electrode (cathode) and a cathode electrode (anode), and a solid polymer electrolyte membrane sandwiched between these electrodes. Usually, both electrodes are formed of a catalyst layer that performs an oxidation / reduction reaction in contact with the solid polymer electrolyte membrane and a gas diffusion layer in contact with the catalyst layer. In the stack 2, when hydrogen is supplied to the anode flow path 21 formed on the anode electrode side and oxygen-containing air is supplied to the cathode flow path 22 formed on the cathode electrode side, these electrochemical reactions occur. To generate electricity.

  The output current taken out from the stack 2 during power generation is input to the battery B and the load (travel motor M, air compressor 41, etc.) via the current controller 29. The ECU 6 calculates a required value for the output current of the stack 2 based on the output signal from the accelerator opening sensor that detects the opening of the accelerator pedal (not shown). The current controller 29 controls the output current of the stack 2 during power generation using the required value calculated by the ECU 6.

  The battery B stores electric power generated by the stack 2 and electric energy recovered as a regenerative braking force by the traveling motor M. Further, for example, when the output current of the stack 2 is limited at the start of the fuel cell system 1 or when the vehicle is operating at a high load, the power stored in the battery B supplements the output of the stack 2. Supplied to the load.

  The anode system 3 includes a hydrogen tank 31 that stores hydrogen gas at a high pressure, a hydrogen supply pipe 32 that extends from the hydrogen tank 31 to the introduction portion of the anode flow path 21 of the stack 2, and a discharge system that discharges the anode flow path 21 to the cathode system 4. A hydrogen discharge pipe 33 extending to a diluter (not shown) provided in FIG. 5 and a hydrogen reflux pipe 34 branched from the hydrogen discharge pipe 33 to the hydrogen supply pipe 32 are configured. The hydrogen circulation flow path for the gas containing hydrogen is constituted by a hydrogen supply pipe 32, an anode flow path 21, a hydrogen discharge pipe 33 and a hydrogen reflux pipe 34.

  In the hydrogen supply pipe 32, in order from the hydrogen tank 31 side to the stack 2 side, a shutoff valve 321 and an injector 322 for injecting new hydrogen gas supplied through the shutoff valve 321 toward the stack 2, An ejector 323 for circulating the gas refluxed from the hydrogen reflux pipe 34 to the stack 2 is provided. The shut-off valve 321 is an electromagnetic valve that opens and closes in response to a command signal from the ECU 6. The injection amount of hydrogen gas from the injector 322 is controlled by PWM control by the ECU 6.

  The hydrogen reflux pipe 34 is provided with a hydrogen pump 341 that pumps gas from the hydrogen discharge pipe 33 side to the hydrogen supply pipe 32 side. In the hydrogen discharge pipe 33, a catch tank 331 for storing water discharged together with the gas from the anode flow path 21 in order from the stack 2 side to the cathode system 4 side, and the gas in the hydrogen circulation flow path are supplied to the cathode system 4. And a purge valve 332 for discharging to the side. The hydrogen pump 341 operates in response to a command signal from the ECU 6. The rotational speed of the hydrogen pump 341 is controlled by the ECU 6. The purge valve 332 is an electromagnetic valve that opens and closes in response to a command signal from the ECU 6.

  The catch tank 331 is provided with a drain pipe 35 for discharging the accumulated water. The drain pipe 35 extends from the catch tank 331 to the downstream side of the purge valve 332 in the hydrogen discharge pipe 33. A drain valve 351 is provided in the drain pipe 35. When the drain valve 351 is opened, the water accumulated in the catch tank 331 is discharged to a diluter (not shown) of the cathode system 4 through the hydrogen discharge pipe 33. The drain valve 351 is an electromagnetic valve that opens and closes in response to a command signal from the ECU 6.

  The cathode system 4 includes an air compressor 41, an air supply pipe 42 extending from the air compressor 41 to the introduction section of the cathode flow path 22, an air discharge pipe 43 extending from the discharge section of the cathode flow path 22 to a diluter (not shown), An air recirculation pipe 45 that branches from the discharge pipe 43 and reaches the air supply pipe 42 and a humidifier 46 that connects the air discharge pipe 43 and the air supply pipe 42 are configured. The oxygen circulation flow path of the gas containing oxygen is constituted by the air supply pipe 42, the cathode flow path 22, the air discharge pipe 43 and the air reflux pipe 45.

  The air compressor 41 supplies outside air to the cathode flow path 22 of the stack 2 through the air supply pipe 42. The air compressor 41 operates in response to a command signal from the ECU 6. The rotational speed of the air compressor 41 is controlled by the ECU 6. The air exhaust pipe 43 is provided with a back pressure valve 432 for adjusting the pressure in the cathode channel 22. The back pressure valve 432 is an electromagnetic valve that opens and closes in response to a command signal from the ECU 6. The pressure in the cathode flow path 22 of the stack 2 during power generation is controlled to an appropriate level according to the power generation state of the stack 2 by adjusting the opening of the back pressure valve 432 while supplying air with the air compressor 41. Is done.

  The humidifier 46 collects water contained in the gas discharged from the cathode channel 22 and humidifies the air supplied from the air compressor 41 using the collected water. With the function of the humidifier 46, the MEA of the stack 2 during power generation is maintained in a wet state suitable for power generation.

  The air supply pipe 42 is provided with a bypass pipe 47 that bypasses the humidifier 46. This bypass pipe 47 is provided with a bypass valve 471. When the bypass valve 471 is opened, most of the air supplied from the air compressor 41 is supplied to the stack 2 via the bypass pipe 47, that is, bypassing the humidifier 46. The bypass valve 471 is an electromagnetic valve that opens and closes in response to a command signal from the ECU 6.

  The air supply pipe 42 and the air discharge pipe 43 are provided with an inlet sealing valve 421 and an outlet sealing valve 431, respectively. When these sealing valves 421 and 431 are closed, the inside of the cathode flow path 22 of the stack 2 is blocked from outside air. These sealing valves 421 and 431 are electromagnetic valves that open and close in response to a command signal from the ECU 6.

  The cooling device 5 includes a refrigerant circulation path 51 including the inside of the stack 2 as a part of the flow path, a water pump 52 provided in the refrigerant circulation path 51 for circulating the refrigerant in the circulation path 51, and the refrigerant circulation path 51. And a radiator 53 as a part. The water pump 52 operates in response to a command signal from the ECU 6. The number of rotations of the water pump 52 is controlled by the ECU 6.

  The cooling device 5 circulates the refrigerant by the water pump 52 and promotes heat exchange between the stack 2 and the refrigerant, and cools the refrigerant by the radiator 53, thereby exceeding the upper limit temperature defined for protecting the stack 2. Do not.

  The ECU 6 includes a plurality of impedance sensors 24, a cathode side temperature sensor 25, an anode side temperature sensor 26, a refrigerant temperature sensor 27, a current sensor 28, and a plurality of pieces for grasping the state of the fuel cell system 1 such as the cell voltage detection device 29. Sensor is connected.

  The impedance sensor 24 detects the impedance resistance value of the stack 2 and transmits a signal substantially proportional to the detected value to the ECU 6. The impedance resistance value of the stack 2 has a correlation with the moisture content of the MEA inside the stack 2. More specifically, the impedance resistance value tends to increase as the water content of MEA decreases. The ECU 6 can indirectly acquire the water content state of the MEA by using the impedance resistance value detected by the impedance sensor 24. The current sensor 28 detects the output current of the stack 2 and transmits a signal substantially proportional to the detected value to the ECU 6.

  The cathode side temperature sensor 25 detects the temperature of the gas discharged from the cathode flow path 22 and transmits a signal substantially proportional to the detected value to the ECU 6. The anode side temperature sensor 26 detects the temperature of the gas discharged from the anode flow path 21 and transmits a signal substantially proportional to the detected value to the ECU 6. The ECU 6 calculates the internal temperature of the stack 2 by using the detection values of these temperature sensors 25 and 26. The refrigerant temperature sensor 27 detects the temperature of the refrigerant discharged from the stack 2 in the circulation path 51 and transmits a signal substantially proportional to the detected value to the ECU 6.

  The cell voltage detection device 29 detects the voltage (cell voltage) of the fuel cells constituting the stack 2. The cell voltage detection device 29 transmits the average cell voltage detected for each single cell (average cell voltage) or the lowest detected cell voltage (lowest cell voltage) to the ECU 6.

  The ECU 6 also changes the output values of various sensors such as the sensors 24 to 29 over time, and the air compressor 41, the bypass valve 471, the back pressure valve 432, the injector 322, the hydrogen pump 341, the purge valve 332, the drain valve 351, and the water. A history recording device 61 is provided for recording operation histories of various devices such as the pump 52 by electromagnetic means.

  The amount of staying water in the anode flow path 21 of the stack 2 and the amount of staying water in the cathode flow path 22 vary depending on the power generation mode of the stack 2. The ECU 6 refers to the history of output values of various sensors recorded in the history recording device 61 and the operation history of various devices, whereby the amount of accumulated water in the anode channel 21 and the retention in the cathode channel 22 are determined. Calculate the amount of water.

  A driver's seat of a vehicle (not shown) is provided with an ignition switch IG that can be operated by the driver to start and stop the fuel cell system 1. The ignition switch IG outputs a start command signal for the fuel cell system 1 to the ECU 6 when turned on by the driver. Upon receiving this start command signal, the ECU 6 starts system start processing (not shown) for starting power generation by the stack, warm-up control for promoting warm-up of the stack (see FIG. 2 described later), and the like. To do. In addition, after the system startup process is completed and power generation by the stack is possible, the ECU 6 selects and selects a power generation condition suitable for the state of the stack according to the procedure described later with reference to FIGS. Power generation control is performed under power generation conditions. The ignition switch IG outputs a stop command signal for the fuel cell system 1 to the ECU 6 when turned off by the driver. The ECU 6 starts a system stop process (not shown) upon receiving this stop command signal.

  FIG. 2 is a flowchart showing a procedure of warm-up control when starting the fuel cell system. This process is executed by the ECU in response to the reception of the start command signal from the ignition switch.

  In S1, ECU performs the driving | operation which accelerates | stimulates the temperature rise accompanying the electric power generation of a stack, and moves to S2. Here, the driving | operation which accelerates | stimulates the temperature rise accompanying the electric power generation of a stack | stump | means reducing the rotation speed of a water pump or stopping rotation, for example. Further, for example, so-called low IV power generation (power generation in a state where the internal resistance of the stack is increased by reducing the amount of air supplied to the stack) that is advantageous for warming up the stack may be performed. In S2, the ECU acquires the temperature of the stack based on the output of the temperature sensor, and proceeds to S3. In S3, the ECU determines whether or not the acquired stack temperature is higher than a warm-up end determination temperature determined to determine the end of warm-up control. If the determination in S3 is NO, the process returns to S1 to continue the warm-up control, and if the determination in S3 is YES, this process ends.

  FIG. 3 is a diagram for explaining the contents of the generation and discharge model of the stagnant water in the stack, which is assumed in the power generation control of the stack according to the present embodiment. The upper time chart of FIG. 3 shows changes in the impedance resistance value of the stack during power generation, the amount of water contained in the MEA, and the load (output current) of the stack. The lower part of FIG. 3 is a diagram schematically showing the state of water in the gas flow path (cathode flow path or anode flow path) inside the stack at each time.

  In the stagnant water model of FIG. 3, the moisture present in the stack is considered separately for those contained in the MEA (membrane water content) and those contained in the gas flow path (stagnant water). The membrane water content is assumed to be generated primarily by power generation, and the stagnant water is the water content of the membrane that has exceeded the water retention limit of the MEA, and the membrane water content changes. This is considered to be generated secondarily. Further, in the stagnant water model of FIG. 3, it is considered that the impedance resistance value of the stack has a direct correlation with the amount of water content of the membrane. Based on such a model, the flow of stagnant water from the generation of the stagnant water in the gas flow path (cathode flow path or anode flow path) in the stack to the discharge from the stack is the load of the stack. 3 (for example, output current) and the impedance resistance value of the stack can be expressed as shown in the lower part of FIG.

  For example, between times t1 and t3, the output current of the stack increases and decreases. The water content of MEA increases and decreases between time t2 and time t4 with a slight delay from the change in the output current of the stack. As described above, since the impedance resistance value has a correlation with the water content of the MEA, such a change in the water content of the MEA is detected as an increase or decrease in the impedance resistance value. Here, in the stagnant water model of FIG. 3, the stagnant water is treated as if the water content of the membrane exceeding the water retention limit exudes. Therefore, the period from time t2 to t4 when the impedance resistance value becomes smaller than the predetermined value is a section in which the accumulated water is newly generated in the gas flow path. The staying water generated between time t2 and time t4 is pushed out together with the gas flowing in the flow path after time t4.

  Thereafter, between time t5 and time t7, the output current of the stack increases and decreases again. As a result, between time t6 and t8, new accumulated water is generated again before the accumulated water generated between time t2 and t4 is discharged out of the stack. And the staying water produced | generated between the time t6-t8 is extruded outside with the gas which transfers to the time t6 transition and the inside of a flow path, and is all discharged | emitted out of a stack at the time t9. As a result, the stagnant water hardly exists in the gas flow path inside the stack.

  Based on the above stagnant water model, it is considered that the stagnant water exudes beyond the water retention limit, so that the amount of stagnant water can be controlled to some extent by controlling the water content.

  FIG. 4 is a flowchart showing a processing procedure for determining the presence or absence of stagnant water based on the stagnant water model shown in FIG. More specifically, the process of FIG. 4 determines whether or not stagnant water exists in the gas flow path (cathode flow path and anode flow path) inside the stack using the above stagnant water model. This is a process of switching the staying water flag indicating the presence or absence of staying water between “0” (no staying water) and “1” (with staying water) according to the result. This process is repeatedly executed by the ECU under a predetermined control period when the activation command signal from the ignition switch is received.

  In S11, the ECU determines whether or not the detected value of impedance resistance is greater than a predetermined stagnant water generation threshold. If the detected value of the impedance resistance is greater than the accumulated water generation threshold, it can be determined that no accumulated water is generated (see S12). If the detected value is less than or equal to the accumulated water generation threshold, it is determined that the accumulated water has occurred. Yes (see S13). The impedance resistance value of the stack decreases as the membrane water content increases, and stagnant water is generated when the membrane water content exceeds the water retention limit. The stagnant water generation threshold value in S11 is a threshold value that is determined in order to determine whether or not stagnant water is generated based on the detected value of the impedance resistance, and is the same as the later-described excessive humidification lower limit value described with reference to FIG. Are set to approximately the same value. Further, this stagnant water generation threshold value may be a fixed value or may be changed according to the internal temperature of the stack.

  If the determination in S11 is NO, the ECU determines that stagnant water has newly occurred (see S13), and resets the value of the drainage counter to a predetermined maximum value (see S14). Here, the “drainage counter” corresponds to the distance from the position where the stagnant water exists in the gas flow path to the outlet position of the gas flow path. The stagnant water can be generated anywhere in the gas flow path. In the process of FIG. 4, it is assumed that the time taken to discharge becomes the longest, and that new accumulated water is always generated on the inlet side of the gas flow path. Therefore, the maximum value in S14 is set to the total length of the gas flow path, for example.

  By the way, the position where the accumulated water is generated and the position where it is not generated may be estimated to some extent based on the specifications of the stack, the internal temperature, and the like. For example, stagnant water is unlikely to be generated in the vicinity of the gas channel inlet, and the region in which stagnant water is unlikely to be generated may expand into the gas channel as the internal temperature of the stack increases. Therefore, the maximum value of S14 may be a fixed value as described above, or the position where the stagnant water is generated may be estimated each time based on the specifications of the stack, the internal temperature, or the like, and may be changed accordingly.

  In S15, the ECU determines the drainage speed based on the stack operating conditions. Here, the “drainage speed” corresponds to the distance that the accumulated water flows along the gas flow path during the control period of FIG. Here, the “stack operating condition” for determining the drainage speed is, for example, the type of the power generation mode currently being executed (for example, as shown in FIG. 5 described later, the drain power generation mode, the normal power generation mode, the dry power generation mode). Mode, wet power generation mode, etc.), and a required value for the output current of the stack, but the present invention is not limited to these. The drainage speed is determined based on a required value for the output current of the stack, using a map determined for each power generation mode as exemplified in FIG. 4, for example. As shown in FIG. 4, the drainage speed is set to a smaller value as the required value becomes smaller. Therefore, the smaller the required value, the longer it takes for the generated accumulated water to be discharged out of the stack. In addition, when the drainage speed is compared between these four power generation modes under the same required value, the drainage speed in the drainage mode is the largest.

  In S16, the value of the drainage counter is updated by subtracting only the drainage speed calculated in S15 from the value of the drainage counter. For convenience of calculation, the minimum value of the drainage counter is 0. In S17, it is determined whether or not the drainage counter is 0 or less. When the drainage counter is 0 or less, it is determined that the amount of staying water in the gas flow path is 0 (that is, there is no staying water in the gas flow path), and the staying water flag is set to clarify this. Set to 0 (see S18). If the drain counter is greater than 0, it is determined that the amount of staying water in the gas flow path is not 0 (that is, stagnant water exists in the gas flow path), and the staying water flag is set to 1 to clearly indicate this. (See S19). It should be noted that the stagnant water flag updated every predetermined control cycle in the process of FIG. 4 is referred to as appropriate in the process described later with reference to FIGS.

  Next, the four power generation modes of the stack defined in this embodiment will be described. The ECU can perform power generation control under four types of power generation conditions having different functions. Here, the power generation conditions refer to control modes of various devices necessary for generating power by the stack. In addition, various devices necessary for power generation include, for example, a humidifier bypass valve 471, an air compressor 41, a back pressure valve 432, a water pump 52, and the like. In the present embodiment, four types of power generation conditions are defined: a reference condition, a wet condition, a dry condition, and a drainage condition. In the following, performing power generation control under reference conditions is referred to as power generation in the normal power generation mode, performing power generation control under wet conditions is referred to as power generation in the wet power generation mode, and performing power generation control under dry conditions. What is performed is referred to as power generation in the dry power generation mode, and power generation control under drainage conditions is also referred to as power generation in the drainage power generation mode.

  FIG. 5 is a table comparing four types of power generation conditions from the functional aspect. More specifically, the amount of change per unit time in the water content of MEA and the amount of stagnant water when power generation is performed constantly under equal environmental conditions (for example, required current value, outside air temperature, etc.) It is the figure compared on the basis of reference conditions.

  By the way, the condition for establishing the accumulated water model described with reference to FIG. 3 is limited to the case where both the moisture content of the stack and the amount of accumulated water are within a predetermined range. For example, when the output current of the stack rapidly decreases, both the moisture content of the stack and the amount of accumulated water may increase in a short time, and in such a case, the above-described accumulated water model may not be established. In the upper and lower stages of FIG. 5, the water content of the stack and the amount of stagnant water are suppressed to such an extent that the stagnant water model of FIG. It is shown separately when there are too many water models.

  As shown in FIG. 5, the wet conditions are determined such that the water content increases more than the reference conditions, the dry conditions, and the drainage conditions. In addition, the wetting conditions are determined so that the amount of stagnant water increases more than the reference conditions, drying conditions, and drainage conditions. The drying conditions are determined so that the water content is reduced more quickly than the reference conditions. Further, the drying conditions are determined so that the amount of the staying water is quickly reduced as compared with the reference conditions.

  When the water content of the stack and the amount of stagnant water are suppressed to such an extent that the stagnant water model shown in FIG. 3 is established, the drainage conditions are determined so that the amount of stagnant water can be reduced more quickly than at least the reference conditions. . Regarding the change in water content, the drainage conditions are determined to be almost the same as the standard conditions.

  When the water content of the stack and the amount of stagnant water are so large that the stagnant water model of FIG. 3 does not hold, the drainage conditions are determined so that the amount of stagnant water quickly becomes smaller than the reference conditions and the drying conditions. Further, the drainage conditions are determined so that the water content is quickly reduced as compared with the standard conditions and the drying conditions.

  Next, an example of a specific control mode of each power generation condition for realizing the function as shown in FIG. 5 will be described.

<Humidifier bypass valve>
The humid condition is determined so that the humidification amount of air by the humidifier is larger than the reference condition. The drying condition is determined so that the humidification amount of air by the humidifier is smaller than the reference condition. Note that the humidification amount of air by the humidifier specifically refers to the amount of water supplied to the air by the humidifier per unit time. Therefore, the opening of the bypass valve of the humidifier is controlled to be closer to the valve closing side when power generation control is performed under humid conditions than when power generation control is performed under reference conditions. When performing control, control is performed so that the valve is opened more than when power generation control is performed under reference conditions. Thereby, during the wet power generation mode, air that is wetter than during the normal power generation mode is supplied to the stack, and during the dry power generation mode, air that is drier than during the normal power generation mode is supplied to the stack. Note that, under dry conditions, it is preferable to fully open the bypass valve of the humidifier and minimize the amount of air humidified by the humidifier.

<About air compressor and back pressure valve>
The wetting condition is determined so that the volume flow rate of the air flowing through the cathode channel of the stack is smaller than the reference condition. On the other hand, the drying conditions are determined so that the volume flow rate is higher than the reference conditions. The drainage conditions are determined so that the volume flow rate is larger than the reference conditions and the drying conditions. Therefore, the supply amount of air newly supplied from the air compressor (that is, the rotation speed of the air compressor) is controlled to be smaller than the reference condition under the wet condition, and is lower than the reference condition under the dry condition. Also, it is controlled to be larger than both the reference condition and the drying condition under the drainage condition. At the same time, the opening of the back pressure valve that adjusts the pressure in the cathode flow path of the stack is controlled to be smaller to the closed side (pressure increasing side) than the reference condition under the wet condition, and under the dry condition. It is controlled to be larger on the open side (decompression side) than the reference condition, and is controlled to be larger on the open side (decompression side) than both the standard condition and the drying condition under the drainage condition.

<About water pump>
The wetting condition is determined so that the circulation amount of the refrigerant is larger than the reference condition. The drying conditions are determined so that the circulation amount of the refrigerant is smaller than the reference conditions. Note that the refrigerant circulation amount specifically refers to the amount of refrigerant per unit time supplied to the stack by driving the water pump. Therefore, when power generation control is performed under wet conditions, the rotation speed of the water pump is set to be higher so that the amount of refrigerant circulating is larger than when power generation control is performed under reference conditions. In addition, when the power generation control is performed under the drying condition, the rotation speed of the water pump is set to be lower so that the circulation amount of the refrigerant is smaller than when the power generation control is performed under the reference condition. In the case where the refrigerant circulation amount is changed for each power generation condition as described above, the ratio between the off time and the on time in the intermittent operation of the water pump may be increased or decreased instead of increasing or decreasing the rotation speed of the water pump. .

  By the way, when an air compressor or a water pump is driven, a slight vibration and noise are generated. Therefore, it is preferable that these rotational speeds are as low as possible regardless of power generation conditions. Therefore, when the functions of the above-described four types of power generation conditions can be achieved only by adjusting the humidification amount of air by the humidifier, the volume flow rate of air, the circulation amount of the refrigerant, etc. are positively applied under these four types of power generation conditions. There is no need to make a difference.

  As described above, the wetting conditions are power generation conditions suitable for quickly moistening the MEA when the water content is low (dry taste). On the contrary, the drying conditions are power generation conditions suitable for quickly drying the MEA when the water content is high (moist). Therefore, by performing power generation in the wet power generation mode or the dry power generation mode at an appropriate timing, the moisture content of the stack can be maintained within a predetermined range suitable for power generation. Hereinafter, for example, controlling the water content by performing power generation in the wet power generation mode or the dry power generation mode at a timing determined based on the detection value of the impedance sensor is referred to as water content control.

  The drainage condition is a power generation condition suitable for quickly discharging out of the stack when there is a large amount of stagnant water. Also, as shown in the lower table of FIG. 5, when the water content of the stack and the amount of stagnant water are so large that the stagnant water model of FIG. Both the amount of water and stagnant water can be quickly discharged out of the stack. Hereinafter, for example, controlling the amount of staying water by performing power generation in the drainage power generation mode at a timing determined based on the amount of staying water calculated by the processing of FIG. 4 is referred to as staying water control.

  FIG. 6 is a flowchart showing a procedure of power generation condition selection processing. This power generation condition selection process is a process for selecting an optimal power generation condition from the four types of conditions having different effects as described above. This process is repeatedly executed by the ECU under a predetermined control cycle in response to the fact that the stack can generate power after receiving the activation command signal from the ignition switch.

  In S21, the ECU refers to the stack current history of the operation history recording device, and calculates the stack current decrease rate (past stack current value−current stack current value / predetermined time) from the present to a predetermined time. . In S22, it is determined whether or not the calculated reduction rate is larger than a predetermined output reduction rate threshold value. For example, the case where the stack current increases corresponds to, for example, a case where the driver requests rapid acceleration. The case where the stack current becomes small corresponds to, for example, a case where the driver requests acceleration and then requests rapid deceleration, or a case where the driver requests cruise travel after acceleration. Immediately after the stack current has once increased and then suddenly decreased, the moisture content control of the stack or the accumulated water control (see FIGS. 7 and 9 to be described later) can catch up with the optimization of the moisture content inside the stack. In many cases, both the MEA and the gas flow path are wet in the stack. The processing of S21 and S22 is processing for estimating that the entire inside of the stack is excessively humidified based on the stack current reduction rate. The specific value of the output decrease rate threshold value in S22 is determined by performing a test in advance for such a purpose.

  When the determination in S22 is NO, the ECU determines whether or not the lowest cell voltage of the stack is equal to or lower than a predetermined threshold (see S23). For some reason, the moisture content control or stagnant water control of the stack (see FIGS. 7 and 9 described later) did not function, resulting in flooding that would cause the inside of the stack to become excessively wet. In some cases, the minimum cell voltage may greatly decrease. The process of S23 is a process of estimating that the entire inside of the stack is excessively humid based on a decrease in the minimum cell voltage of the stack. The specific value of the threshold for the lowest cell voltage in S23 is determined by conducting a test in advance for such a purpose.

  When the determinations at S22 and S23 are both NO, the ECU executes the process of FIG. 7 for selecting an appropriate power generation condition with the water content control being higher than the staying water control. On the other hand, if any of the determinations in S22 and S23 is YES, the ECU performs the stay water control rather than the water content control in order to quickly discharge the moisture inside the stack that is excessively humid. The process of FIG. 9 for selecting an appropriate power generation condition as the upper level is executed.

  FIG. 7 is a flowchart showing a procedure of power generation condition selection processing when water content control is prioritized. As shown in FIG. 7, when the moisture content control is prioritized, the moisture content control that attempts to keep the detected value of the impedance resistance within a predetermined appropriate range is the retained water control for limiting the amount of retained water to a predetermined value or less. Is set higher.

  In S31, the ECU determines whether the impedance resistance value is within an appropriate range specified by the predetermined overhumidification lower limit value and the overdrying upper limit value, is less than the overhumidification lower limit value, or is greater than the overdrying upper limit value. Determine if. When the impedance resistance value is equal to or lower than the overhumidification lower limit value, it is determined that the stack is in an overhumidified state. If the stack during power generation becomes excessively humidified, power generation may become unstable or the stack may deteriorate. Further, when the impedance resistance value is larger than the overdrying upper limit value, it is determined that the stack is in an overdrying state. If the stack during power generation is in an excessively humid state, the stack may deteriorate or the output may decrease. Therefore, it is preferable that the impedance resistance value during power generation is within the above-described appropriate range as much as possible. Note that the excessive humidification lower limit value and the excessive drying upper limit value may be fixed values as with the above-described accumulated water generation threshold value, or may be changed according to the internal temperature of the stack.

  If it is determined in S31 that the stack is in an overdried state, the ECU selects the wet condition as the power generation condition for the stack (see S32). As described with reference to FIG. 5, under the wet condition, the moisture content of the stack changes to the increasing side, so that the impedance resistance value decreases toward an appropriate range.

  When it is determined in S31 that the stack is in an excessively humidified state, the ECU proceeds to S33, determines whether or not the warm-up control (see FIG. 2) has been completed, and the warm-up control has been completed. Is selected as the power generation condition of the stack (see S34). As described with reference to FIG. 5, under the dry condition, the moisture content of the stack changes to the decreasing side, so that the impedance resistance value increases toward an appropriate range. If it is determined in S34 that the warm-up control has not been completed, the ECU selects the drainage condition as the stack power generation condition, as will be described later, regardless of the determination result in S31 (see S36). .

  When it is determined in S31 that the impedance resistance value is within the appropriate range, the ECU determines whether or not the stagnant water flag is 1 (see S35). When it is determined in S35 that the stagnant water flag is 1, that is, when it is estimated that the stagnant water remains in the gas flow path, the ECU sets the drainage condition as the power generation condition of the stack. Select (see S36). As described with reference to FIG. 5, the drainage condition has a higher drainage rate than the other power generation conditions, so that the accumulated water is quickly discharged under the drainage condition. If it is determined in S35 that the stagnant water flag is 0, the ECU selects the reference condition as the power generation condition for the stack (see S37).

  FIG. 8 is a time chart showing an example when the power generation condition is selected under the process of FIG. An example of power generation mode transition realized by the processing of FIG. 7 will be described with reference to FIG. FIG. 8 shows a case where the stack current temporarily increases or decreases from time t1 to t3 and from time t5 to t7, as in FIG.

  As shown in FIG. 8, when the stack current increases at time t1, the MEA water content increases and the impedance resistance value decreases slightly later. At time t2, it is determined that the state is an overhumidified state in response to the impedance resistance value falling below the overhumidified lower limit value (see S31 in FIG. 7). As a result, the power generation mode becomes the dry power generation mode set so as to promote the drying of the MEA until it is determined that the impedance resistance value again exceeds the overhumidification lower limit value and is not in the overhumidity state at time t4 (dry power generation). Process). Further, at time t2, the drainage counter is reset to the maximum value in response to the impedance resistance value falling below the stagnant water determination threshold value (see S11 in FIG. 4) set to substantially the same value as the overhumidification lower limit value. (See S14 in FIG. 4). The drain counter is repeatedly reset to the maximum value while the impedance resistance value is below the stagnant water determination threshold (see S14 in FIG. 4).

  Thereafter, at time t4, the impedance resistance value exceeds the humidification lower limit value and the stagnant water determination threshold value. That is, at time t4, the water content falls within a range suitable for power generation, and generation of new accumulated water stops. For this reason, at time t4, the dry power generation mode for promoting the drying of the MEA is completed (see S31 in FIG. 7). At time t4, subtraction of the drainage counter is started in response to the stop of the generation of new accumulated water (see S16 in FIG. 4). In addition, after time t4, the water content of the membrane falls within the appropriate range, but the remaining water remains in the gas flow path of the stack. Until the accumulated water in the gas flow path is discharged, more specifically, until the drainage counter becomes 0 (see S17 in FIG. 4), the power generation mode becomes the drainage power generation mode (see S36 in FIG. 7).

  Thereafter, in response to the increase in the stack current at time t5, the impedance resistance value again falls below the excessive humidification lower limit value and the staying water determination threshold value at time t6. As a result, similarly to the time t2, the power generation mode shifts from the drainage power generation mode to the dry power generation mode, and the drainage counter is reset to the maximum value.

  At time t8, the impedance resistance value exceeds the over-humidification lower limit value and the stagnant water determination threshold value. As a result, similarly to time t4, the power generation mode shifts from the dry power generation mode to the drainage power generation mode, and the subtraction of the drainage counter starts. After time t8, the state in which the impedance resistance value exceeds the excessive humidification lower limit value and the stagnant water determination threshold value is maintained, whereby the drainage counter gradually decreases and becomes 0 at time t9. As a result, the drainage power generation mode ends at time t9, and thereafter, the power generation mode becomes the normal power generation mode. As described above, according to the processing of FIG. 7, after the dry power generation mode is completed, power generation is performed in the drainage power generation mode until a waiting time until the drainage counter becomes zero has elapsed.

  FIG. 9 is a flowchart showing a procedure of power generation condition selection processing when stagnant water control is prioritized. As shown in FIG. 9, at the time of staying water control priority, in order to give priority to the positive discharge of the staying water that is estimated to be accumulated, the staying water control is opposite to the water content control, contrary to the process of FIG. Set higher.

  In S51, the ECU determines whether or not the staying water flag is 1. When it is determined that the staying water flag is 1, the ECU selects the drainage condition as the power generation condition of the stack regardless of the detected value of the impedance resistance (see S52). Thereby, discharge of the staying water accumulated in the gas flow path is promoted. In addition, when the process of FIG. 9 is executed, both the water content and the amount of accumulated water are often excessive. Under such circumstances, as described with reference to the lower part of FIG. 5, it is possible to reduce the water content of the membrane in a larger amount than in the draining condition than in the drying condition. Therefore, in the process of FIG. 9, the retained water control is placed above the moisture content control, and the drainage condition is selected as the power generation condition, so that the retained water is discharged quickly and at the same time the membrane water content is an appropriate amount. Can be reduced quickly.

  In S53, the ECU determines whether the impedance resistance value is within an appropriate range, is less than the excessive humidification lower limit value, or is greater than the upper limit upper drying value. When it is determined in S53 that the stack is in an overdried state (that is, when the impedance resistance value is larger than the overdrying upper limit value), the ECU selects a wet condition as a power generation condition to eliminate the overdried state (S54). reference). If it is determined in S53 that the impedance resistance value is within the appropriate range, the ECU selects the reference condition as the power generation condition for the stack (see S57).

  When it is determined in S53 that the stack is in an excessively humidified state (that is, when the impedance resistance value is smaller than the excessively humidified lower limit value), the ECU determines whether or not the warm-up control has been completed (see S55). If the warm-up control is completed, the drying condition is selected as the power generation condition (see S56). If it is determined in S55 that the warm-up control is not completed, the ECU selects the drainage condition as the stack power generation condition regardless of the determination result in S53.

  Although one embodiment of the present invention has been described above, the present invention is not limited to this. Within the scope of the gist of the present invention, the detailed configuration may be changed as appropriate.

1 ... Fuel cell system 2 ... Stack (fuel cell)
3. Cathode system (oxidant gas supply device)
4 ... Anode system (fuel gas supply device)
24. Impedance sensor (impedance detection device)

Claims (5)

A fuel cell that generates electricity when supplied with oxidant gas and fuel gas;
An oxidant gas supply device for supplying an oxidant gas to the fuel cell;
A fuel gas supply device for supplying fuel gas to the fuel cell;
An impedance detection device that detects the impedance of the fuel cell, and a control method for a fuel cell system comprising:
For power generation of the fuel cell,
A normal power generation mode in which power generation is controlled under predetermined reference power generation conditions;
A dry power generation mode for performing power generation control under a dry power generation condition determined so that the membrane of the fuel cell is dried more than the reference power generation condition;
Three power generation modes: a stagnant water discharge power generation mode in which power generation control is performed under a discharge power generation condition that is set so that more stagnant water in the gas flow path of the fuel cell is discharged than the reference power generation condition Is defined,
An overhumidification determination step of determining whether or not the fuel cell is in an overhumidified state based on a detection value of the impedance detection device;
A dry power generation step of generating power in the dry power generation mode until it is determined that the over-humidification state is not over-humidified by the over-humidification determination step;
A request acquisition step of acquiring a required value for the output of the fuel cell;
After the dry power generation process is completed, power generation is performed in the stagnant water discharge power generation mode until a state in which the detection value of the impedance detection device exceeds a predetermined lower limit threshold is maintained for a waiting time determined according to the acquired request value. A method for controlling a fuel cell system, comprising: Prior SL accumulated water discharge step, the control method as set forth in claim 1, characterized in that lengthening of the waiting time as the required value the acquired decreases. An output reduction rate determination step for determining whether the output reduction rate of the fuel cell is greater than a predetermined output reduction rate threshold;
And a drainage priority step of generating power in the accumulated water discharge power generation mode regardless of the determination result of the overhumidification determination step when the decrease rate is greater than the output decrease rate threshold. Item 3. A method for controlling a fuel cell system according to Item 1 or 2. A voltage drop determination step for determining whether or not the voltage of the fuel cell is equal to or lower than a predetermined voltage threshold;
2. A drainage priority step of generating power in the stagnant water discharge power generation mode regardless of the determination result of the overhumidification determination step when the voltage is equal to or lower than the voltage threshold value. Or a control method of the fuel cell system according to 2; Further comprising a warm-up determination step of determining whether or not the fuel cell has been warmed-up,
4. The fuel cell system according to claim 1, wherein, when the warm-up is not completed, the stagnant water discharge step is performed regardless of a determination result of the overhumidification determination step. 5. Control method.

Images