JPS63241873A - Air flow controller for fuel cell - Google Patents

Abstract

PURPOSE:To control a plant automatically and stably by introducing a controller for which target values for parameters under control and their time differentiations are inputted and to which operational control laws can be applied as they are instead of a PI controller. CONSTITUTION:If a direct current IDC is given, a target value for an interpole differential pressure DPP corresponding to the current is calculated from a memory 22. Respective target values of an output command ACW, a set value of DC voltage VDC, a target value for an interpole differential pressure DPP, and a set value for quantity of hydrogen at the outlet of an air pole CGH are compared with outputs ACW', VDC', DPP', and CGH' being quantities of present states of the actual plant by adders 21A-21D, and their feedback deviations EAC, EVD, EPP, and ECH are calculated, where EAC and EVD are added 23, so a deviation of electricity quantity EDP is obtained. Each deviation is differentiated by differentiators 24A-24C. These deviations and differentiations are inputted to controllers 25A-25C, and obtained control outputs dFI1-dFI3 are sent to changeable gain proportioning devices 24A-26C and an adder 27, so a weighted mean control output dFI is obtained.

Description

[Detailed description of the invention] [Structure of the invention] (Industrial application field) The present invention relates to an air flow rate control device for a fuel cell.

(Prior Art) Electric power is generated by rotating a generator, usually a prime mover such as a steam turbine, and converting the applied mechanical energy into electrical energy in the generator. on the other hand,
Steam that drives a steam turbine or the like is generated using thermal energy obtained by burning fuel such as oil or gas in a boiler or the like. The method of extracting the chemical energy of this fuel as thermal energy, converting it into the energy of steam, and then converting mechanical energy such as a steam turbine into electrical energy is disadvantageous in terms of efficiency, so in recent years, Fuel cell power generation, which directly converts chemical energy into electrical energy, has come to be adopted as an energy-saving power generation method. This fuel cell generates electricity by chemically changing the supplied fuel, but the output is direct current, so if it is consumed as is in a specific area, it is consumed as direct current, or it is converted to alternating current by a DC/AC converter to generate electricity. Connected to the grid.

FIG. 4 is an explanatory diagram of a typical water-cooled fuel cell plant that uses hydrocarbons as raw fuel. In the figure, the part indicated by a dotted line a is the fuel cell plant 1.

The raw fuel enters the mixer 14 with its flow rate controlled by the raw fuel control valve 10. On the other hand, the mixer 14 receives steam whose flow rate is controlled by the steam control valve 16 from the steam generator I5. The raw fuel mixed with steam in the mixer 14 enters the reformer 4, where it is heated and reformed into hydrogen. The reformed fuel is then passed through a high temperature shift converter 8. It passes through the low-temperature shift converter 9 and becomes reformed fuel with a higher hydrogen content. The reformed fuel enters the fuel electrode 5A of the fuel cell 5 with its flow rate controlled by the reformed fuel control valve 11, where a part of the hydrogen is consumed as an electrical energy source, and the rest is sent to the main burner 12 of the reformer 4. The gas is combusted and becomes high-temperature gas for heating the reformer 4, which joins the exhaust air from the air electrode 5B of the fuel cell 5, passes through the combustor 7, enters the turbine 2 of the turbo compressor, and drives the compressor 3 connected thereto.

The flow rate of the air discharged from the compressor 3 is controlled by the air control valve 13 and enters the air electrode 5B of the fuel cell 5.

A part of the oxygen in the air that has entered the air electrode 5B reacts with hydrogen in the fuel electrode 5A, and the rest is discharged from the air electrode 5B as exhaust air, joining with the exhaust air from the main burner 12 of the reformer 4. It is used to drive the turbine 2 of the turbo compressor via the combustor 7.

The fuel cell 5 is cooled by a battery cooling water circulation pump 19 that supplies battery cooling water to the steam generator 15 and to the cooling plate 5C.

It is circulated to the steam generator 15. This temperature control of the battery cooling water is performed by adjusting the amount of cooling of the steam generated in the steam generator 15 in the heat exchanger 20 using the battery cooling water temperature control valve 17.

The fuel cell 5 generates electrical energy through a catalytic reaction between hydrogen at the fuel electrode 5A and oxygen at the air electrode 5B so that the air electrode 5B becomes a positive electrode and the fuel electrode 5A becomes a negative electrode. Supply that electrical energy to the load.

The hydrogen and oxygen respectively supplied to the inlets of both poles react in approximately proportion to the electrical energy generated at that time and become water, and the unreacted portion is discharged from each scraping port.

Normally, the DC output of the fuel cell 5 is converted into AC by a converter 6 and sent to the power system.

The above is the basic configuration and general operation of the fuel cell plant.

FIG. 5 shows a conventional air flow rate control device, which calculates a pre-correction air flow rate command 38 corresponding to a direct current 35 of a fuel cell. On the other hand, the converter actual output 31 is fed back to the output command 3° to take the deviation, and the DC voltage 32 is fed back to the DC voltage setting 33 to take the deviation, and the sum of the 1-car deviations is controlled by the PI amount. An output deviation correction 34 is calculated, and a correction 37 is determined by a limit calculation based on a correction amount limit 36 corresponding to the DC current 35. Further, a correction 37 is added to the pre-correction air flow rate command 38 to calculate an air flow rate command 39, and an actual air flow rate 40 is fed back to this, and the air control valve 13 is controlled by the PI amount based on this deviation. .

(Problem to be Solved by the Invention) Generally, the differential pressure between the fuel electrode and the air electrode of a fuel cell is determined by the difference in pressure drop from each pole to the confluence point of the main burner exhaust and exhaust air of the reformer. The flow rates of reformate and air that occur and thus flow through each pole are closely related to the interpole differential pressure. Furthermore, since the electrolyte layer between the fuel electrode and the air electrode in a fuel cell has very low strength, it is necessary to always keep the pressure difference between the electrodes small.

Conventionally, one way to keep this differential pressure between poles small is to
The operator was manually adjusting the air flow rate.

Furthermore, in the chemical reaction of a fuel cell, if the amount of oxygen at the air electrode, that is, the flow rate of air is insufficient, a phenomenon occurs in which hydrogen is generated in the amount of air at the same time as the DC voltage decreases. A hydrogen amount detector is installed at the air electrode outlet, and when the hydrogen amount increases, the operator makes manual adjustments such as increasing the air flow rate.

As described above, the conventional air flow rate control device has a problem in that an operator must manually adjust the flow rate depending on the plant situation.

[Structure of the invention]

(Means for solving problems) An air flow control device for a fuel cell according to the present invention.

Since adjusting the air flow rate is effective for adjusting the differential pressure between the electrodes and hydrogen generation at the air electrode, the deviation between the output command and the actual output, the DC voltage, and the air flow rate command according to the DC current are effective. , air flow rate command with flow rate correction corresponding to the deviation between the setting and the actual DC voltage, the deviation between the inter-electrode differential pressure setting and the actual differential pressure, and the deviation between the hydrogen amount setting at the air electrode outlet and the actual hydrogen amount. It is characterized by automatically providing stable plant operation as a result of controlling the flow rate.

Furthermore, the present invention is characterized by the introduction of a controller that inputs the target value of the controlled parameter and its time derivative instead of the PI quantity controller, and allows the control law used by the operator to be applied as is. .

(Function) The object of the present invention is to provide an air flow control device for a fuel cell that can automatically and stably control a plant from a low load state to a rated output state.

(Example) Hereinafter, an example of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of an air flow control device for a fuel cell according to an embodiment of the present invention. In FIG. 1, reference numeral 22 denotes a schedule memory for the inter-electrode differential pressure, which outputs the inter-electrode differential pressure target DPP when the DC current IDC is input. 21 is output command A
Deviation between CW and actual output AcW′ EAC=, ACW−AC
W', DC voltage setting VDC and actual DC voltage VDC:
' Deviation between EVD = VDC - VDC', deviation between target differential pressure D P P and actual differential pressure DPP' EPP = DP
P-DPP', an adder that calculates the deviation ECH=CGH-CGH' between the air electrode outlet hydrogen amount setting CGH and the actual hydrogen amount CGH'; 23 is the sum ED of the deviation EAC and EVD;
P=EAC+EVD (hereinafter referred to as EDP and EPP).

24 is the time differential of the deviation dEDP=dEDP/
dt, dEpp=dEPP/dt, dECH=dECH
/dt (hereinafter, when dEDP, dEPP, and dECH are collectively referred to, they will be referred to as Δe). Reference numeral 25 denotes a controller which inputs the deviation of each weight and the differential value of the deviation and calculates each control output ΔU based on the following control laws 1 to 5.

Control law 1: The deviation e is large in the jJE direction, and the differential value Δ of the deviation
When e is large in the negative direction, the control output ΔU is made small in the positive direction.

Control law 2: When the deviation e is large in the positive direction and the differential value Δe of the deviation is large in the positive direction, the control output ΔU is made large in the positive direction.

Control law 3: When the deviation e is close to zero, the control output ΔU is set to a value close to zero, regardless of the value of the differential value Δe of the deviation.

Control law 4: When the deviation e is large in the negative direction and the differential value Δe of the deviation is large in the negative direction, the control output ΔU is made large in the negative direction.

Control law 5: When the deviation e is large in the negative direction and the differential value Δe of the deviation is large in the positive direction, the control output ΔU is made small in the negative direction.

FIG. 2 is a diagram illustrating a specific calculation method of this controller 25. In the same figure, each graph has a deviation e on the horizontal axis,
The differential value Δe of the deviation and the control output ΔU are -100 to 10.
0%, and the vertical axis represents the concepts of "large in the positive direction", "small in the positive direction", "zero", "small in the negative direction", and "large in the negative direction" as a set. Measure μ to which each concept belongs
is taken from 0 to 1, and each control law described above is expressed.

That is, looking at (control law 1) in the same figure, in order to calculate the measure μe from the deviation e at this time, the deviation e is defined as large from 10% to 100% in the positive direction, and the measure μe is calculated as the deviation A pattern p in which the measure μe is set to 0 when is 10%, increases gradually as the deviation e increases, reaches a maximum of 1 at a deviation of 70%, and then decreases again to 0 at a deviation of 100%.
We have a g engineering facility. Here, the reason why the deviation is set to be the highest at 70% and the measure is decreased thereafter is that the deviation e is 7
The maximum value is around 0%, which means that deviations larger than that are unlikely to occur under normal control conditions.

Next, in order to calculate the measure μΔe from the differential value Δe of the deviation, the differential value Δe of the deviation is defined as small from -10% to -100%, and the measure μΔe is.

A pattern PΔ8□ is provided in which the maximum value is 1 at −60%.

Further, the control output ΔU is defined as being small in the positive direction from −20% to +70%, and a pattern PΔU1 is provided in which the measure μΔU becomes the highest 1 at +20%.

These patterns P. As can be seen from pΔ0 and PΔU, in control law 1, the control deviation e is large in the positive direction, but the differential value Δe of that deviation is large in the negative direction, which means that the deviation is quickly heading toward recovery. in case of.

The control output ΔU is kept small and correction operations are performed sparingly.

This means preventing excessive control in the opposite direction.

Further, the magnitude of the control output ΔU at this time is determined according to the magnitude of the deviation e and the differential value Δe of the deviation.

Hereinafter, one illustrated pattern P is used for control laws 2 to 5 in the same manner. 2~Pos, PΔo2~PΔes+PΔU2~P
ΔU5 is provided.

Each controller 25 is equipped with the control laws 1 to 5, and first calculates the deviation e inputted therein and the differential value Δe of the deviation based on those control laws, and first calculates the measure μe obtained from each pattern of e and Δe.
, μΔe, cut off the upper part of the pattern of control output ΔU using the smaller μM I N, find the remaining portion PBΔU for each control law, calculate their maximum value μMAX, and calculate the average value of the erroneous pattern PμMAXΔU. The output of each controller 25 is dFIi.

For example, when the values of e=40% and Δe=30% are input, in control law 1, μ, 1=0.7. μΔo1=0, μM
I N, = 0 In control law 2, μ. 2=0.7. μΔ. ,
=0.5, μM I N, =0.5 control law 3 /j
ei”o−2yμΔo3=0.2, μMIN3=0
．． 2 In control law 4, μQ4”OtμΔe4=0, μMI
In N4=02 control law 5, +uos"0-2y μΔes
”, μM I N, = 0, and only control law 2 and control law 3 can be applied. Regarding these control laws, P
BΔU was taken from the shaded area PBΔU2 in Figure 2.
and PBΔU3. The maximum value μMAX for these PBΔU□ and PBΔU was calculated at the shaded area P in Figure 2.
From this average value, the output dF of the controller 25 is μMAXΔU.
Calculating Ii 6 Reference numeral 26 in FIG. 1 is a variable gain proportional device, and the gain is changed by direct current.

Figure 3 shows an example. The gain for the output and DC voltage is large at low currents, the gain for the electrode differential pressure is 2, and the gain for the air electrode outlet hydrogen amount is 3.
is large at intermediate to high currents.

Finally, 27 is the output, DC voltage, and differential pressure between poles.

This is an adder that adds the outputs of each controller related to the amount of hydrogen at the air electrode outlet. This output corresponds to the conventional air flow rate correction and is used in place of A in Figure 5. With the above configuration, the DC current IDC is given. An inter-electrode differential pressure target value DPP corresponding to the current is calculated from the inter-electrode differential pressure schedule memory 22. The target values of the output command ACW, the DC voltage setting VDC, the above-mentioned electrode differential pressure target value DPP, and the air electrode outlet hydrogen amount setting CGH are calculated by adding 121A, 21
At B, 21C, and 21D, the output ACW', the DC voltage VDC', the interelectrode differential pressure DPP', and the daily hydrogen amount CGH', which are the current state quantities of the actual plant, are compared and the feedback deviation EAC=ACW is compared. -ACW'
, EVD=VDC-VDC', PPP=DPP-D
PP', ECH=CGH-CGH' are calculated, and the output deviation EAC and voltage deviation EVD are added by an adder 23 to calculate the electric quantity deviation EDP=EAC+EVD. In addition, differentiators 24A, 24B, and 24G calculate the time differentiation of each deviation dEDP=dEDP/dt, dEPP=dEPP/
dt and dEcH=dEcH/dt are calculated. This deviation and the differential value of the deviation are sent to the controller 25A as e=EDP, Δe=dEDP, e=EPP, Δe=dE
To the controller 25B as P P, e = E CH, Δe
=dECH to the controller 25C, each control output dFI,, d
FI,, dFI, are obtained.

That is, regarding the output and DC voltage, as mentioned above, input e=EDP, Δe=dEDP,
Each pattern P shown in FIG. ,~po,,pΔ0tech~P
Δ0. from each measure μ. 1~μOS+ μΔei
~μΔ. , is calculated. Furthermore, the value μ of each of those cars is
Find MIN□~μMIN, and from them calculate the base pattern PBΔU1~ of the controller cover turn PΔU□~PΔUs.
PBΔU is obtained. Furthermore, the maximum value pattern PμMAXΔU obtained by combining these base patterns PBΔU-PBΔU5 is calculated, and the average value of this pattern, that is, the weighted average value of the control output ΔU spread over a certain range, is calculated to calculate the final control output. dF11 is obtained.

Each controller 25A, 25B obtained in this way.

Control output d F I,, d F 1. from 25C. ,d
F I □ variable gain proportional device 26A, 26B,
26G and passes through the adder 27 to obtain the weighted average value dF.
I becomes clFI=□Xd F 1. +, Xd F I, +
, it is calculated as Xd F L. Therefore, this dF
I is input to the limiter in place of the output of the broken line block A shown in FIG. 5, and the obtained output is used as the pre-correction air flow rate command 3.
By using the air flow rate command 39 in addition to the air flow rate command 39 and controlling the air control valve 13, it is possible to automatically and stably control plant state quantities such as output, DC voltage, interelectrode differential pressure, and daily hydrogen amount of air discharge. become able to.

〔Effect of the invention〕

As described above, according to the present invention, in addition to the output and DC voltage, the interelectrode pressure difference and the daily hydrogen amount of the air are taken into consideration as correction amounts when calculating the air flow rate command from the DC current. Even if the output, DC voltage, interelectrode differential pressure, and air discharge daily hydrogen amount fluctuate due to operational disturbances, stable plant operation can be achieved by automatically adjusting the air flow rate. Further, unlike conventional PID controllers, hunting due to saturation of the differential integrator does not occur even if a large deviation occurs, and stable and quick control operation 7 can be achieved.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram of an air flow control device for a fuel cell according to an embodiment of the present invention, FIG. 2 is an explanatory diagram of the operation of the controller of the air flow control device of FIG. 1, and FIG. 3 is a diagram similar to the one shown in FIG. Figure 4 is an explanatory diagram of a general water-cooled fuel cell, and Figure 5 is a block diagram of a general air flow control apparatus for the fuel cell shown in Figure 4. It is a diagram. 21... Adder. 22...Schedule memory for differential pressure between poles. 23, 29...adder, 24...differentiator. 25:...Controller, 26...Proportional device. Agent Patent Attorney Noriyuki Chika Yudo Hirofumi MitsumataFigure 3Figure 4

Claims (2)

[Claims] (1) In a fuel cell air flow control device that controls the air flow rate supplied to the fuel cell air electrode according to the output current of the fuel cell, a means for calculating at least a deviation between an output command and an actual output; and a DC voltage. A means for calculating the deviation between the set value and the actual DC voltage, a means for calculating the deviation between the set value of the interelectrode differential pressure and the actual differential pressure, and a means for calculating the deviation between the set value of the hydrogen amount at the air electrode outlet and the actual hydrogen amount. means for calculating, means for calculating a correction amount for the air flow rate command based on each of the deviations, means for calculating the air flow rate command by adding the correction amount to the air flow rate command for the output current; 1. An air flow control device for a fuel cell, comprising means for adjusting an actual air flow rate in accordance with a flow rate command. (2) In a fuel cell air flow control device that controls the air flow rate supplied to the fuel cell air electrode according to the output current of the fuel cell, a means for calculating at least a deviation between an output command and an actual output; and a DC voltage. means for calculating the deviation between the set value and the actual DC voltage; means for calculating the deviation between the electrode differential pressure set value and the actual differential pressure; and means for calculating the deviation between the hydrogen amount set value at the air electrode outlet. , a means for time-differentiating each of the deviations, inputting each of the deviations and their time derivatives, and specifying in advance "greater in the positive direction", "smaller in the positive direction", "zero", and "negative". direction small” and “
The operator's concept of ``greater in the negative direction'' is given as a measurement parameter, and the control output is further set as ``larger in the positive direction'', ``smaller in the positive direction'', ``zero'', and ``in the negative direction'' for combinations of the above concepts. A measure parameter is also given to the operator's control law that makes it "small" and "large in the negative direction," and when each deviation and its time derivative are input, the corresponding measure of the deviation and its time derivative is calculated for each control law. Select the minimum value, set the measure distribution of the control output of each control law to be valid only for those below the minimum value, and recalculate the bottom degree distribution.
After superimposing the measurement distributions of each control output obtained as a result so as to select the maximum value, the control means determines the control output by averaging, and the output of each of these control means is multiplied by a coefficient and added. means for calculating a correction amount for an air flow rate command; means for calculating an air flow rate command by adding the correction amount to an air flow rate command corresponding to an output current; and means for adjusting an actual air flow rate in accordance with the air flow rate command. An air flow control device for a fuel cell, characterized by comprising: