Fuel cell stack state determination device

Abstract

Provided are a fuel cell condition determination device and a fuel cell vehicle, wherein a computation load is minimized and an internal condition of a fuel cell can be accurately measured in a short time. This fuel cell condition determination device uses membrane resistance, charge transfer resistance, and diffusion resistance in a relationship between measurement frequency and impedance as reference values to generate a load waveform for measurement that will be used the current time, said relationship having been generated on the basis of an existing measurement load waveform that was used the previous time or a waveform for training, applies, to a cell, a waveform having an amplitude and a frequency that were calculated on the basis of the load waveform for measurement, and causes a current or a voltage corresponding to the frequency of this load waveform for measurement to be discharged from the cell, to measure the AC impedance and perform a condition determination including a deterioration diagnosis.

Description

[Technical Field]

[0001] The present disclosure relates to a state determination device for, for example, a fuel cell stack, and a fuel cell vehicle equipped with this state determination device. [Background technology]

[0002] Transportation is essential in modern society, and various vehicles, including automobiles, travel the roads in our daily lives. Among these, fuel cells, which have a relatively low environmental impact, are attracting attention as a new power source for supplying driving force to vehicles.

[0003] In such fuel cells, fuel gas (hydrogen) is supplied to one electrode (fuel electrode) and oxidant gas (oxygen) is supplied to the other electrode (air electrode), and electrical energy is generated through a chemical reaction between these. Therefore, in order to continuously obtain appropriate electrical energy (generated power) from the fuel cell, it is necessary to appropriately determine the deterioration of the fuel cell installed in the vehicle.

[0004] For example, in Patent Document 1 listed below, activation overvoltage is calculated based on the amount of electricity generated during reduction of the catalyst in each cell that makes up the fuel cell, and the FC voltage is estimated from the calculated activation overvoltage. Patent Document 1 proposes comparing the estimated FC voltage with the actual FC voltage detected by a voltage sensor, and determining fuel cell degradation based on the comparison result.

[0005] Furthermore, for example, Patent Document 2 focuses on the water content in a fuel cell as a criterion for determining deterioration, and proposes estimating the water content by determining the proton transfer resistance and gas reaction resistance using a Cole-Cole plot, which is a characteristic diagram showing the relationship between frequency and impedance on a complex plane. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] WO2016 / 067430 issue [Patent Document 2] Japanese Patent Application Laid-Open No. 2007-012418 [Patent Document 3] Japanese Patent Application Laid-Open No. 2012-078339 [Patent Document 4] Japanese Patent Application Laid-Open No. 2007-250365 [Patent Document 5] Japanese Patent Application Laid-Open No. 2001-277056 Summary of the Invention [Problem to be solved by the invention]

[0007] Not limited to the above-mentioned patent documents, current technologies still do not meet market needs, and the following problems remain. Specifically, the AC impedance method, as shown in the aforementioned patent documents, is commonly used as a method for estimating the internal state of a fuel cell during operation. While this AC impedance method is known to use a frequency response analyzer (FRA), it typically uses a single frequency as the measurement waveform, and analyzing multiple frequencies requires time. On the other hand, there is also a method for quickly determining the impedance spectrum using the fast Fourier transform (FFT) method with a waveform in which multiple frequencies are superimposed. However, noise-induced errors occur, which necessitates noise countermeasures and waveform processing, resulting in a large-scale device.

[0008] To perform such high-precision measurements in a short time, it is preferable to use a waveform in which multiple frequencies are superimposed. However, simply implementing the above-mentioned FFT method in a vehicle such as a fuel cell vehicle would require an increase in the scale of the equipment, which is not realistic from a cost perspective.

[0009] The present disclosure has been made in consideration of the above-mentioned problems as an example, and aims to provide a fuel cell state determination device and a fuel cell vehicle that can reduce the computational load and accurately measure the internal state of a fuel cell in a short period of time. [Means for solving the problem]

[0010] In order to solve the above problem, according to one aspect of the present disclosure, there is provided a fuel cell stack state determination device including a fuel cell stack composed of one or more cells, and a control device that applies a measurement load waveform to the cells to determine the state of the fuel cell stack, wherein the control device uses as reference values ​​membrane resistance, charge transfer resistance, and diffusion resistance in the relationship between measurement frequency and impedance generated based on a previously used measured load waveform or a learning waveform, generates a measurement load waveform to be used this time, applies to the cells a waveform having an amplitude and multiple frequencies calculated based on the measurement load waveform, and discharges a current or voltage corresponding to the frequency of the measurement load waveform from the cells to measure AC impedance and determine the state. [Effects of the Invention]

[0011] According to the present disclosure, it is possible to reduce the calculation load and accurately measure the internal state of a fuel cell in a short period of time. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a schematic diagram showing an example of the configuration of a fuel cell vehicle equipped with a fuel cell state determination device according to an embodiment; [Figure 2] FIG. 2 is a schematic diagram illustrating the configurations and functions of a fuel cell vehicle according to an embodiment. [Figure 3] FIG. 2 is a schematic diagram illustrating an example of the configuration of the periphery of a control device that also functions as a state determination device according to an embodiment. [Figure 4] 1 is a flowchart showing a management method including state determination of a fuel cell according to an embodiment. [Figure 5] FIG. 10 is a schematic diagram showing an example of a primary process learning waveform calculated in the fuel cell management method. [Figure 6] 10 is a first example of a graph showing the relationship between frequency and impedance used to obtain the initial value of membrane resistance Rmem in the fuel cell management method. [Figure 7]10 is an example of a second graph defining the relationship between phase difference and frequency used in the measurement frequency exploration step in the fuel cell management method. [Figure 8] FIG. 10 is a schematic diagram showing an example of a waveform for secondary process learning calculated in the fuel cell management method. [Figure 9] 10 is a third example of a graph showing the relationship between frequency and impedance used to obtain the initial value of membrane resistance Rmem in the fuel cell management method. [Figure 10] 10 is an example of a fourth graph defining the relationship between phase difference and frequency used in the measurement frequency exploration step in the fuel cell management method. [Figure 11] FIG. 10 is a schematic diagram showing an example of a measurement load waveform calculated in the fuel cell management method. [Figure 12] FIG. 1 is a schematic diagram showing an example of a Nyquist diagram (Cole-Cole plot) in a fuel cell management method. [Figure 13] FIG. 10 is a schematic diagram showing an example of a current value map used in a learning waveform generation step in the fuel cell management method. [Figure 14] 1 is a schematic diagram showing an example of a water content determination map used in a state determination step in a fuel cell management method. FIG. [Figure 15] FIG. 10 is a schematic diagram showing an example of reference table data used in a water content adjusting step in a fuel cell management method. DETAILED DESCRIPTION OF THE INVENTION

[0013] Next, preferred embodiments of the present disclosure will be described. In this specification and drawings, components having substantially the same functional configurations are designated by the same reference numerals, and redundant description will be omitted. Furthermore, configurations other than those described in detail below may be supplemented appropriately with, for example, the known AC impedance method, impedance measurement techniques disclosed in Patent Documents 1 to 5, or known elemental technologies and configurations related to fuel cell systems and fuel cell vehicles.

[0014] <Fuel cell vehicle FCV> 1 and 2 are schematic diagrams respectively showing an example of the configuration and functional blocks of a fuel cell vehicle FCV equipped with a fuel cell stack FC according to this embodiment. As shown in Fig. 2, this fuel cell vehicle FCV is configured as a four-wheel drive vehicle in which drive torque output from a drive power source 21 that generates drive torque for the vehicle is transmitted to a left front wheel 3LF, a right front wheel 3RF, a left rear wheel 3LR, and a right rear wheel 3RR (hereinafter collectively referred to as "wheels 3" unless a distinction is required). In this embodiment, the drive power source 21 can be, for example, a known electric motor arranged on the front wheel side.

[0015] The electric motors serving as driving force source 21 in this embodiment may be arranged one on each of the front and rear wheels, or one electric motor may be arranged for each wheel 3. In addition to the electric motor described above, driving force source 21 may also include an internal combustion engine such as a gasoline engine, a diesel engine, or a gas turbine engine.

[0016] A power supply system that supplies desired power to such driving force source 21 includes a fuel cell stack FC configured by stacking multiple known fuel cell cells (hereinafter simply referred to as "cells") such as PEFCs (polymer electrolyte fuel cells), a hydrogen gas supply unit including a known hydrogen tank 23 and piping, an air supply unit including a known compressor 31 and piping, a known secondary battery 50 such as a lithium ion secondary battery or a lead storage battery, a known converter 22, and a control device 100 that controls these. Note that the control device 100 in this embodiment also functions as a state determination device 10 that determines deterioration of the fuel cell stack FC. In this power supply system, the fuel cell stack FC and the secondary battery 50 are each capable of supplying power to a load including the electric motor described above.

[0017] As shown in Fig. 2, this fuel cell stack FC is connected to a load including the above-mentioned converter 22 and driving power source 21 (electric motor). Also as shown in Fig. 2, the current and voltage in the fuel cell stack FC are detected by a known current sensor SR1 and voltage sensor SR2, respectively.

[0018] The converter 22 is configured to include a known AC / DC converter that converts DC current to AC current, and a known DC / DC converter that adjusts the voltage of DC current to a desired voltage. As an example, the converter 22 of this embodiment has functions such as receiving a control signal from the control device 100 to set the output voltage that is generated and output by the fuel cell stack FC, and boosting the power generated by the fuel cell stack FC to a desired voltage when supplying it to a load.

[0019] In addition, the fuel cell vehicle FCV of this embodiment is equipped with the above-mentioned driving force source 21, electric steering device 8, and brake devices 4LF, 4RF, 4LR, and 4RR (hereinafter collectively referred to as "brake device 4" unless a distinction is required) as equipment used for driving control.

[0020] The driving force source 21 outputs a driving torque that is transmitted to the front drive shaft 2F and the rear drive shaft 2R via a transmission, a front wheel differential mechanism 5F, and a rear wheel differential mechanism 5R (not shown). The driving of the driving force source 21 and the transmission is controlled by one or more electronic control units (ECUs). The control is performed by a known control device including:

[0021] The front wheel drive shaft 2F is provided with an electric steering device 8. The electric steering device 8 includes an electric motor and a gear mechanism (not shown), and is controlled by a vehicle drive control device 20 (described later) to adjust the steering angles of the left front wheel 3LF and the right front wheel 3RF.

[0022] The vehicle drive control device 20 includes one or more known electronic control units (ECUs) that control the drive of a drive power source 21 that outputs drive torque for the fuel cell vehicle FCV, an electric steering device 8 that controls the steering wheel 9 or the steering angle of the steering wheels, and a brake device 4 that controls the braking force of the fuel cell vehicle FCV. The vehicle drive control device 20 may also have a function of controlling the drive of a transmission that changes the speed of the output from the drive power source 21 and transmits it to the wheels 3.

[0023] Also, as shown in FIG. 2, in the hydrogen gas supply section for supplying fuel (hydrogen) gas to the fuel cell stack FC, the hydrogen gas stored in the hydrogen tank 23 is supplied to the anode side flow path of the above-mentioned fuel cell stack FC via a hydrogen intake valve 32a having a known structure installed in the hydrogen supply flow path.

[0024] A portion of the hydrogen gas discharged from the fuel cell stack FC may be returned to the hydrogen supply flow path by a circulation flow path and a known circulation pump 45. The remainder of the hydrogen gas discharged from the fuel cell stack FC is diluted by a diluter 41 at predetermined timing under the control of the control device 100 via the opening and closing operation of a known hydrogen exhaust valve 32b, and then released (exhausted) to the atmosphere.

[0025] 2, the air supply unit for supplying oxygen gas (air) to the fuel cell stack FC is configured to include, in addition to the above-mentioned compressor 31, a known oxygen intake valve 32c and an air discharge valve (back pressure valve) 32d that adjust the amount of oxygen (air) supplied to the cell. The air supply unit may also include a known flow rate sensor (not shown) that can measure the flow rate of air supplied to the fuel cell stack FC.

[0026] The air taken in by the compressor 31 is supplied to the cathode-side flow path in the fuel cell stack FC via the oxygen intake valve 32c and a known humidifier (not shown). The air supplied to the cells is also supplied to the diluter 41 as cathode off-gas under the control of the air exhaust valve (back pressure valve) 32d by the control device 100.

[0027] The control device 100 includes one or more processors (CPUs (Central Processing Units) ) and one or more memories communicably connected to the one or more processors. The control device 100 may be configured to be connectable to a known external network NET such as the Internet via various known communication devices CD, for example, a form using a smartphone.

[0028] Such a control device 100 may be connected directly or via a CAN (Controller Area Network) or The compressor 31 and each valve 32 are connected via a communication means such as LIN (Local Internet). (hydrogen intake valve 32a, hydrogen exhaust valve 32b, oxygen intake valve 32c, and oxygen exhaust valve 32d), well-known sensors SR such as a current sensor SR1 and a voltage sensor SR2 are electrically connected.

[0029] The fuel cell stack FC of this embodiment has a stack structure in which a plurality of known cells, each having an electromotive force of, for example, about 1 V, are connected in series and stacked. As an example, the fuel cell stack FC of this embodiment can be a polymer electrolyte fuel cell (PEFC) having a structure in which a plurality of cells are connected in series within a pair of known end plates that pressurize and hold the fuel cells at both ends so as to provide the system voltage required by a fuel cell vehicle FCV.

[0030] Each cell constituting the fuel cell stack FC has a structure in which a known MEA (membrane electrode assembly) is interposed between a pair of known separators installed on the fuel electrode side and the air electrode side, respectively. This MEA is configured to include at least a known cathode catalyst layer, a known anode catalyst layer arranged opposite the cathode catalyst layer, and a known polymer electrolyte membrane arranged between the cathode catalyst layer and the anode catalyst layer. Note that the membrane electrode assembly may also be configured to include a known air electrode-side gas diffusion layer and a known fuel electrode-side gas diffusion layer.

[0031] <State Determination Device 10> Next, a state determination device 10 that determines the state of the fuel cell stack FC in this embodiment will be described with reference to Figure 3. That is, the state determination device 10 of this embodiment is configured to have the function of determining the state of the fuel cell stack by applying a measurement load waveform to the cells that make up the fuel cell stack FC.

[0032] The state determination device 10 includes a load waveform generation unit 10A, a current measurement unit 10B, a voltage measurement unit 10C, an impedance measurement unit 10D, and a determination unit 10E. As described above, the state determination device 10 is configured as one function executed by the control device 100 of this embodiment. As shown in FIG. 3 , the control device 100 may include a drive control unit 30 and a presentation control unit 40.

[0033] The load waveform generating unit 10A calculates the membrane resistance R in the relationship between the measurement frequency and the impedance, which is generated based on the previously measured load waveform or the learning waveform used last time. mem , charge transfer resistance R ct , and diffusion resistance R dif as a reference value to generate a learning waveform to be compared with the measurement load waveform. Furthermore, the load waveform generating unit 10A of this embodiment is configured to have a function of generating a measurement load waveform in which a plurality of frequencies having amplitudes and frequencies are superimposed, based on the learning waveform.

[0034] The current measurement unit 10B is configured to have the function of measuring the current value of the fuel cell stack FC described above. More specifically, the current measurement unit 10B of this embodiment can measure the value of the current flowing through the fuel cell stack FC via the current sensor SR1 described above.

[0035] The voltage measurement unit 10C is configured to have the function of measuring the voltage value of the fuel cell stack FC described above. More specifically, the voltage measurement unit 10C of this embodiment can measure the voltage value applied to the fuel cell stack FC via the voltage sensor SR2 described above.

[0036] The impedance measurement unit 10D is configured to have the function of applying a measurement AC signal to the fuel cell stack FC to measure the AC impedance of the fuel cell stack FC. The impedance measurement unit 10D can measure the impedance of the fuel cell stack FC by the AC impedance method described above, based on the current and voltage values ​​measured by applying a current based on the measurement load waveform generated by the load waveform generation unit 10A described above, for example.

[0037] The determination unit 10E has a function of determining the state of the cells constituting the fuel cell stack FC based on the characteristics of the measured AC impedance. More specifically, the determination unit 10E of this embodiment measures the AC impedance by discharging a current or voltage corresponding to the frequency of the measurement load waveform from the cell, and determines the state.

[0038] An example of a state determination that the determination unit 10E can perform is determining whether the amount of water in the fuel cell stack is appropriate. That is, the determination unit 10E of this embodiment determines the membrane resistance R of the cell based on, for example, a Nyquist diagram (Cole-Cole plot) obtained from the AC impedance. mem , charge transfer resistance R ct and diffusion resistance R dif As will be described later, the determination unit 10E does not necessarily need to calculate a Cole-Cole plot, but can instead calculate the membrane resistance R in the relationship between the measurement frequency and the impedance generated based on the measured load waveform. mem , charge transfer resistance R ct , and diffusion resistance R dif may be used as a reference value.

[0039] An example of a state determination that the determination unit 10E can perform is a degradation determination as to whether the fuel cell stack has deteriorated. That is, as will be described later, the determination unit 10E of this embodiment may determine that degradation has occurred in the fuel cell stack FC if the voltage is equal to or lower than a threshold value even after the water management optimization process that has been executed based on the determination of the water content state.

[0040] The drive control unit 30 is configured to have the function of controlling the drive of the fuel cell stack FC. More specifically, when the determination unit 10E determines that degradation has occurred in the fuel cell stack FC, the drive control unit 30 may execute control in accordance with this degradation. More specifically, the drive control unit 30 of this embodiment can execute processing in accordance with the degradation level of the fuel cell stack FC, based on the degree of voltage drop relative to the threshold value.

[0041] As an example, the drive control unit 30 can classify the degree of deterioration into, for example, three levels and then execute processing (such as drive control of the fuel cell vehicle FCV) according to the level. For example, when the degree of deterioration (the degree of voltage drop based on the threshold value) is level 3, which is high, the drive control unit 30 may limit the output from the fuel cell stack FC or may execute a process to prompt a user such as a driver to perform maintenance immediately via the presentation device DD.

[0042] Also, for example, when the level of deterioration is a medium level 2, the drive control unit 30 may immediately execute a warning process via the presentation device DD to notify the driver that deterioration of the fuel cell stack FC is occurring. Furthermore, for example, when the level of deterioration is a relatively low level 1, the drive control unit 30 may execute a warning via the presentation device DD to notify the driver that deterioration of the fuel cell stack FC is progressing after the fuel cell vehicle FCV has been used, for example, when the system has stopped.

[0043] The presentation control unit 40 executes a process of presenting various information such as the deterioration state of the fuel cell stack FC and warnings via a presentation device DD including a known in-vehicle speaker SP and display DP. The presentation control unit 40 may present the various information to the occupant via the in-vehicle presentation device DD, or may access an external terminal such as a smartphone carried by the occupant and perform control to present the information.

[0044] <Fuel cell stack management method> Next, a fuel cell stack management method that can be executed by the control device 100 including the state determination device 10 of this embodiment will be described with reference to Figures 4 to 12. Note that this management method may be used as an algorithm of a computer-readable program. A program having such an algorithm may be distributed, for example, via a known network so that it can be downloaded to a fuel cell vehicle FCV, or may be distributed in the form of being stored on a recording medium. The following description will be given taking as an example the case where a user activates the system of a fuel cell vehicle FCV and starts driving.

[0045] First, in step 1, the control device 100 determines whether or not a deterioration determination of the fuel cell stack FC is necessary. The control device 100 may, for example, base the timing for starting such a deterioration determination on whether or not the fuel cell stack FC has reached an appropriate temperature based on a known temperature sensor (not shown). Furthermore, the control device 100 may, for example, base the timing on when the temperature of the fuel cell stack FC has reached an appropriate temperature and the load on the fuel cell stack FC is relatively small, such as when the fuel cell vehicle FCV is traveling at a constant speed.

[0046] If it becomes necessary to determine the deterioration of the fuel cell stack FC in step 1, the control device 100 determines in the following step 2A whether waveform learning is necessary for the measurement load waveform to be applied to the cells constituting the fuel cell stack FC. For example, when the fuel cell stack FC is started for the first time after the system of the fuel cell vehicle FCV is started, the control device 100 does not yet hold a reference value for the current start-up period, so the control device 100 determines that waveform learning is necessary and proceeds to step 2B. Also, even if the reference value is already held, the control device 100 may determine that waveform learning is necessary if a certain time has passed since the reference value was acquired or if the charge transfer resistance R ct and diffusion resistance R dif In step 2A, it may be determined that waveform learning is necessary even if the learning requirement conditions are met, such as when the reference value is not calculated. On the other hand, if the reference value is held during the current startup period and the learning requirement conditions are not met, the control device 100 determines that waveform learning is not necessary and proceeds to step 3.

[0047] <Learning waveform generation processing> Next, the learning waveform generation process in step 2B will be described. As described above, during initial startup, the control device 100 does not yet retain reference values ​​for the current startup period, and therefore the water content of the fuel cell stack FC is unknown. Therefore, as a first process (measurement frequency search step), the control device 100 measures the impedance of the cells constituting the fuel cell stack FC using a first learning waveform (referred to as a first processing learning waveform) composed of a plurality of roughly selected frequencies. As an example, as shown in FIG. 5, the control device 100 superimposes a first processing learning waveform composed of five frequencies f1 to f5, each with a different number of digits, on a DC current to be applied to the cells to measure the impedance. In this embodiment, for example, the frequency f1 is set to 1 Hz, the frequency f2 to 10 Hz, the frequency f3 to 100 Hz, the frequency f4 to 1 kHz, and the frequency f5 to 10 kHz are set. However, the frequencies constituting the first processing learning waveform may be set to optimal values ​​and numbers of frequencies through experiments or simulations.

[0048] Furthermore, as shown in Fig. 5, the maximum amplitude of the current in the primary processing learning waveform can be set in advance. As an example, in this embodiment, the maximum amplitude of the current value at each frequency of the primary processing learning waveform is set to 0.25 A. Note that the maximum amplitude of the current value may be determined by experiment or simulation, and is preferably any value between 0.01 A and 0.3 A, for example.

[0049] The control device 100 then applies the first processing learning waveform to the cell to perform impedance measurement. The control device 100 also calculates, from the measurement values ​​obtained based on this first processing learning waveform, an impedance-frequency graph (first graph) showing the relationship between impedance and frequency shown in Fig. 6 (although the accuracy is rougher than that of the third graph described later), and a phase difference-frequency graph (second graph) showing the relationship between the phase difference between current and voltage and frequency (although the accuracy is rougher than that of the fourth graph described later).

[0050] The reason why this first treatment is performed in this embodiment is that the membrane resistance R mem This is to roughly grasp the value of . That is, although it is known that the membrane resistance value of a typical cell is generally in the range of 3 kHz to 7 kHz, it is possible that the membrane resistance value of the cell is abnormal and does not fall within the above range due to some factor, such as deterioration or the external environment. Therefore, in this embodiment, the first process is to quickly calculate whether the cell for which deterioration is to be determined is the above-mentioned abnormal cell (in other words, whether the membrane resistance value is in the normally expected range of 3 kHz to 7 kHz) using a relatively low frequency while suppressing the calculation load.

[0051] In the example of the first graph in Figure 6, it can be estimated that the minimum value appears in the region generally above 3 kHz, and furthermore, from the second graph in Figure 7, it can be assumed that the zero cross point appears in the region generally above 3 kHz. Therefore, the control device 100 can determine that the cell being measured this time is not an abnormal cell and proceed to the second processing.

[0052] After the first process in the learning waveform generation process, the control device 100 performs the second process (initial R mem As an acquisition step, the control device 100 performs a process of generating a second-order processing learning waveform by increasing the number of frequencies from each reference frequency, using the multiple frequencies (f1 to f5) used in the first-order processing learning waveform as reference frequencies. At this time, the control device 100 may increase the number of frequencies used for measurement by equally dividing the measurement time by each reference frequency.

[0053] As an example, as illustrated in FIG. 8, the control device 100 can use 1, 2, ..., 9 Hz for frequency f1 (i.e., a total of nine frequencies in increments of 1 Hz from 1 to 9 Hz), 10, 20, ..., 90 Hz for frequency f2 (i.e., a total of nine frequencies in increments of 10 Hz from 10 to 90 Hz), 100, 200, ..., 900 Hz for frequency f3 (i.e., a total of nine frequencies in increments of 100 Hz from 100 to 900 Hz), 1k, 2k, ..., 9 kHz for frequency f4 (i.e., a total of nine frequencies in increments of 1 kHz from 1k to 9 kHz), and 10k, 20k, ..., 90 kHz for frequency f5 (i.e., a total of nine frequencies in increments of 10 kHz from 10k to 90 kHz). In this case, in order to prevent unnecessary load from being applied to the cell, the maximum amplitude of the current in the second-order processing learning waveform is preferably, for example, about 5 A in absolute value. From this perspective, the current amplitude value at each frequency in the second-order processing learning waveform in this embodiment is set to 0.25 A.

[0054] Then, from the measurement values ​​obtained based on this second processing learning waveform, the control device 100 calculates an impedance-frequency graph (third graph) showing the relationship between impedance and frequency shown in Fig. 9, and a phase difference-frequency graph (fourth graph) showing the relationship between the phase difference between current and voltage and frequency shown in Fig. 10. For example, as is clear from a comparison between Fig. 6 and Fig. 9, the number of frequencies used for measurement in the second processing learning waveform is increased compared to the first processing learning waveform. Therefore, the control device 100 of this embodiment calculates the membrane resistance R of the cell based on these third and fourth graphs. mem In this embodiment, the control device 100 measures the membrane resistance R from the minimum value K of the impedance-frequency graph shown in FIG. mem It can be determined that the resistance is 40 mΩ.

[0055] This embodiment is characterized by determining the maximum measurement frequency (maximum measurement frequency) within a range where measurement accuracy can be ensured. That is, as is clear from the Nyquist diagram described later, among the frequencies measured in the cell impedance, the membrane resistance R mem For this reason, in this embodiment, this maximum measurement frequency is determined in advance, and frequencies higher than this are not used for deterioration determination because they have little effect on measurement accuracy, thereby making it possible to reduce the calculation load on the control device.

[0056] In this manner, in this embodiment, the membrane resistance R in the cell is calculated using the frequency-impedance graph (third graph) illustrated in FIG. mem That is, when the above-mentioned second-order processing learning waveform is applied to the cell, a plurality of plots appear according to the above-mentioned plurality of frequencies, and the control device 100 creates a frequency-impedance graph based on the plots of several points, and calculates the minimum value of the graph as the current membrane resistance R of the cell. mem It is determined that:

[0057] In this embodiment, the upper limit of the measurement frequency used to determine cell degradation is determined using the phase difference-frequency graph (fourth graph) shown in Fig. 10. That is, the control device 100 determines the frequency at which the phase difference is 0 (zero) in the phase difference-frequency graph shown in Fig. 10 as the maximum measurement frequency, and determines that frequencies equal to or higher than this maximum measurement frequency will not be used for measurement.

[0058] In the secondary treatment, the membrane resistance R mem After the initial value of and the maximum measurement frequency are determined, the control device 100 calculates the membrane resistance R as the third process (load waveform generation step). mem , charge transfer resistance R ct , and diffusion resistance R difThe control device 100 performs a process of generating a measurement load waveform by increasing the number of frequencies from each reference frequency using the frequencies of the membrane resistance R as a reference frequency. In this case, the control device 100 may increase the number of measurement frequencies by equally dividing the measurement time by each reference frequency. As an example, as shown in FIG. 11, the control device 100 mem The charge transfer resistance R ct 610, 620, 690 Hz (i.e., a total of nine frequencies, increasing by 10 Hz from 610 to 690 Hz), corresponding to the frequency (650 Hz), and the diffusion resistance R dif Corresponding to the frequency (5 Hz), 1, 2, ..., 9 Hz (i.e., a total of nine frequencies can be used, increasing in 1 Hz increments from 1 to 9 Hz). In this case, the maximum amplitude of the current in the measurement load waveform is preferably, for example, about 5 A in absolute value, from the viewpoint of suppressing unnecessary load on the cell.

[0059] After the measurement load waveform is generated in step 2B, the load waveform to be finally applied to the cell is determined in step 3. That is, if the measurement load waveform is generated in step 2B, it is determined that this measurement load waveform will be used as the load waveform in step 3. On the other hand, if the process moves from step 2A to step 3, it is determined that the load waveform used previously will be used.

[0060] Next, in step 4, the control device 100 may perform offset processing on the measurement load waveform as necessary. For example, when the current value of the fuel cell stack FC when performing degradation determination is 0 A, if the width of the DC component applied to the fuel cell stack FC is smaller than the amplitude width of the current, measurement accuracy cannot be guaranteed. For this reason, the control device 100 may perform processing to increase the amplitude width of the current in the measurement load waveform when, for example, the current value of the fuel cell stack FC is measured near 0 A. On the other hand, by specifying the minimum current value of the fuel cell stack FC as the start condition for the deterioration determination, for example, this step 4 can be omitted as appropriate.

[0061] In the following step 5, the control device 100 performs measurement and analysis to determine deterioration using the measurement load waveform. More specifically, the control device 100 measures the impedance obtained from the cell again by superimposing the measurement load waveform obtained in step 3 on the DC current applied to the cell. As a result, the control device 100 measures the membrane resistance R mem , charge transfer resistance R ct , and diffusion resistance R dif are measured respectively. More specifically, the control device 100 calculates the membrane resistance R mem For the above, the control device 100 calculates the real impedance value under a predetermined condition in the relationship between the measurement frequency and the impedance (for example, the zero cross point where the phase difference becomes zero in the phase difference-frequency graph). Note that instead of the above, the control device 100 may calculate the membrane resistance R in the cell based on, for example, a Nyquist diagram (Cole-Cole plot) shown in FIG. mem may be calculated.

[0062] On the other hand, the control device 100 has a charge transfer resistance R ct and diffusion resistance R dif For the above, it is possible to calculate the imaginary impedance value under other predetermined conditions in the relationship between the measurement frequency and the impedance. Note that the above-mentioned "other predetermined conditions" can be exemplified by the point where the slope in the Nyquist diagram (Cole-Cole plot) shown in FIG. 12 becomes negative. That is, the control device 100 calculates the imaginary impedance value at the point where the slope in the Nyquist diagram (Cole-Cole plot) shown in FIG. 12 becomes negative from the charge transfer resistance R ct and diffusion resistance R dif The point where the slope changes from negative to positive is determined as the inflection point P. Generally, the charge transfer resistance R ct The diffusion resistance R difSince the frequency is higher than the inflection point P, the control device 100 calculates the charge transfer resistance R ct and diffusion resistance R dif The values ​​of can be clearly distinguished. In this embodiment, the charge transfer resistance R ct and diffusion resistance R dif Although the imaginary impedance value described above is used as the reference value for the above, the present invention is not limited to this, and for example, the measured frequency may be used as the reference value.

[0063] In step 5, the membrane resistance R mem , charge transfer resistance R ct , and diffusion resistance R dif When each of these values ​​is measured, the control device 100 stores in the memory device MD the frequency of the measurement load waveform used in the measurement along with the current value of the fuel cell stack FC at the time of measurement. This makes it possible to store in the memory device MD the frequency of the measurement load waveform corresponding to the current value applied to the fuel cell stack FC as a current map. Therefore, by measuring the current value of the fuel cell stack FC at the time of measurement for degradation determination, the control device 100 can use the frequency of the measurement load waveform that was optimal in a previous measurement according to that current value for subsequent degradation determinations.

[0064] An example of the current map is shown in Figure 13. In Figure 13, V1 represents the known OCV overvoltage, V2 represents the known resistance overvoltage, V3 represents the known activation overvoltage, V4 represents the known diffusion overvoltage, V5 represents the known anode overvoltage, and V6 represents the cell output voltage. As can be seen from the figure, in this embodiment, the current value applied to the fuel cell stack FC is divided into zones, and the frequency of the measurement load waveform determined in steps 2A to 3 is stored corresponding to each zone. As an example, in this embodiment, the above-mentioned current map can be divided into zones every 10 A. Specifically, if the deterioration determination of this embodiment is performed when the current value applied to the fuel cell stack FC is 35 A, for example, frequency information of the measurement load waveform used in this measurement is stored in the storage device MD, an external server, etc., corresponding to zone 4 to which 35 A belongs.

[0065] Furthermore, the frequency information of the measurement load waveform in each zone may be updated each time a deterioration determination is performed. This allows the latest frequency information of the measurement load waveform to be stored for each zone in response to changes over time in the fuel cell stack FC. Furthermore, in this embodiment, the zones are divided into 10 A zones, but this is not limiting and the number of amperes to be divided into zones may be adjusted depending on the number of amperes that the fuel cell system can output.

[0066] Here, since a learning waveform has already been generated after step 2B, for example, it is possible to execute the following process in step 3 after it has been determined in step 2A that waveform learning is not necessary. That is, when executing a measurement for determining deterioration, the control device 100 measures the current value of the fuel cell stack FC, and can refer to a current map stored in the memory device MD or the like to read out frequency information of a measurement load waveform corresponding to the current value of the fuel cell stack FC at the time of the previous measurement.

[0067] The current map may also be referenced when generating the learning waveform and the measurement load waveform in step 2B. Furthermore, if the sum of the current amplitude values ​​when generating the measurement load waveform exceeds the preset maximum amplitude (5A in absolute value in this example as described above), the number of frequencies in each frequency band f1 to f5 may be thinned out. At this time, by referring to the current map, for example, the number of frequencies in the range of the diffused resistor R difIf the measurement is performed between zones 1 and 3 where noise is unlikely to occur, the number of frequencies in the high frequency bands of frequencies f4 and f5 may be thinned out. In this way, the control device 100 may refer to the current map and perform processing to thin out the number of frequencies in the measurement load waveform based on the current value of the fuel cell stack FC at the time of measurement. This makes it possible to further reduce the processing load on the control device without significantly reducing the accuracy of the measurement.

[0068] The above process results in the cell membrane resistance R mem , charge transfer resistance R ct , and diffusion resistance R dif After each of the membrane resistances R 1 and R 2 is calculated, the control device 100 analyzes the water content state in the cell. More specifically, the control device 100 calculates the membrane resistance R 1 and R 2 shown in FIG. m em and the water content, and the diffusion resistance R dif Based on the relationship between the temperature and the water content, the water content state in the cell can be analyzed. That is, in step 5, the control device 100 calculates the previously acquired parameters (membrane resistance R mem and diffusion resistance R dif ) value and the newly calculated parameter (membrane resistance R mem and diffusion resistance R dif ) is compared with the value of . Based on the result of the comparison, the control device 100 calculates how the newly acquired parameters have changed (transitioned) relative to the already acquired parameters.

[0069] The control device 100 then refers to reference table data in which transition predictions of the water content corresponding to the respective transitions of the membrane resistance and the diffusion resistance are defined, as exemplified in FIG. 15, and executes the water content adjustment process of the cell, as exemplified below. Note that such reference table data may be stored in the above-mentioned storage device MD or an external server outside the vehicle. That is, as shown in the figure, the control device 100 calculates (a) the membrane resistance R mem increases, and the diffusion resistance R d ifIf the membrane resistance R mem decreases and the diffusion resistance R dif If the resistance R increases, it is determined that there is a strong tendency for the cell to become wet, and in the subsequent step 6, a drying process is performed on the cell by a known method. mem increases, and the diffusion resistance R di f If the resistance R decreases, it is determined that the cell is in a humid state, and in the subsequent step 6, a drying process is intermittently performed on the cell by a known method. mem decreases and the diffusion resistance R dif If the value of the cell density also decreases, it is analyzed as being in a drying trend, and in the subsequent step 6, a wetting treatment is intermittently performed on the cell by a known method.

[0070] After adjusting the water content of the cells of the fuel cell stack FC in step 6, the control device 100 executes a degradation diagnosis process for the fuel cell stack FC in the following step 7. That is, in step 7, the control device 100 measures the current value and voltage value of the fuel cell stack FC using, for example, current sensor SR1 and voltage sensor SR2, and determines whether the voltage of the fuel cell stack FC is below a threshold value even after the water content adjustment of the cells in step 6 (for example, if the IV characteristics of the cells are measured in advance and the IV characteristics of the cells at the time of this degradation determination are below a predetermined reference range).

[0071] If the control device 100 determines in step 8A that the fuel cell stack FC has deteriorated, it executes a warning process to a user such as a driver in the following step 8B. Note that the control device 100 may execute a process according to the level of deterioration described above in the warning process.

[0072] That is, when the deterioration level is "1," the control device 100 may, for example, perform the above-described process of calling the attention of the user via the presentation device DD. Also, when the deterioration level is, for example, "2," the control device 100 may, for example, perform the above-described process of presenting the maintenance recommendation to the user via the presentation device DD. Also, when the deterioration level is, for example, "3," the control device 100 may limit the output of the vehicle via the above-described drive control unit 30, or may perform the above-described guidance process for the user via the presentation device DD. In this way, the control device 100 of this embodiment can perform processing according to the deterioration level of the fuel cell stack FC, based on the degree of voltage drop relative to the threshold value.

[0073] The control device 100 including the state determination device 10 of the present embodiment described above can execute a process of varying the frequency and amplitude of the measurement load waveform to be measured this time by feeding back the results of previously measured cell impedance measurements. This makes it possible to maintain the accuracy of measurements, including degradation determination, while reducing the calculation load, even when the cell state changes.

[0074] Although the preferred embodiments of the present disclosure have been described in detail above with reference to the accompanying drawings, the technology of the present disclosure is not limited to such examples. It is clear that a person skilled in the art of the technology to which the present disclosure pertains can conceive of various modified or altered examples within the scope of the technical ideas described in the claims, and it is understood that these also naturally fall within the technical scope of the present disclosure. [Explanation of symbols]

[0075] 10. Status determination device 20 Vehicle drive control device 30 Drive control unit 40 Presentation control unit FC fuel cell stack FCV fuel cell vehicle

Claims

1. a fuel cell stack composed of one or more cells; a control device that applies a measurement load waveform to the cells to determine the state of the fuel cell stack, The control device A learning waveform is generated by referring to the measurement results obtained by applying to the cell a first processing learning waveform, each of which has frequencies with different digits in order to primarily grasp the membrane resistance of the cell, and a second processing learning waveform, in which the number of frequencies used in measurement with the first processing learning waveform is increased in order to secondarily grasp the membrane resistance; a measurement load waveform to be used this time is generated by using as reference values ​​the membrane resistance, charge transfer resistance, and diffusion resistance in the relationship between the measurement frequency and impedance generated based on the learning waveform; applying a waveform having an amplitude and a plurality of frequencies calculated based on the measurement load waveform to the cell; a current or voltage corresponding to the frequency of the measurement load waveform is discharged from the cell to measure AC impedance, thereby determining the state. A fuel cell stack status determination device.

2. The control device determining a water content state of the fuel cell stack by calculating a difference between a previous measurement and a current measurement using the membrane resistance, the charge transfer resistance, and the diffusion resistance; For the film resistance, the real impedance value at the zero cross point in the relationship between the measurement frequency and impedance is used as a reference value, For the charge transfer resistance and the diffusion resistance, an imaginary impedance value at a point where the slope of the relationship between the measurement frequency and the impedance becomes negative is used as a reference value. The fuel cell stack state determination device according to claim 1 .

3. The control device If the voltage remains below a threshold even after the water management optimization process is executed based on the water content determination, it is determined that deterioration has occurred in the fuel cell stack. The fuel cell stack state determination device according to claim 2 .

4. The control device performing a process according to the deterioration level of the fuel cell stack based on the degree of voltage drop relative to the threshold value; The fuel cell stack state determination device according to claim 3 .

Images