• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    In order to be able to regulate the influencing variables of the operation of a fuel cell accurately and quickly, it is provided that a controller (R) is used for the regulation, which was designed on the basis of a model of the gas conditioning unit (3) in the form of a coupled, nonlinear multivariable system by the decoupled and linearized coupled nonlinear multivariable system using Lie derivatives and the controller (R) was designed for the decoupled, linear multivariable system, the controller (R) at each sampling time of the control using predetermined setpoints (yj, dmd) the manipulated variables (uG , uS, uN, Q) for at least three existing actuators of the influencing variables of the gas conditioning unit (3) and at least these three actuators of the gas conditioning unit (3) set the calculated manipulated variables (uG, uS, uN, Q) at each sampling instant of the control ,
    公开号:AT518518A4
    申请号:T50663/2016
    申请日:2016-07-20
    公开日:2017-11-15
    发明作者:Ing Christoph Kügele Dipl；Dr Jakubek Stefan；Kancsár János
    申请人:Avl List Gmbh；
    IPC主号:
    专利说明:

    Controlled gas conditioning for a reaction gas of a fuel cell
    The subject invention relates to a controlled gas conditioning for a reaction gas of a fuel cell and a method for controlling a gas conditioning for a fuel cell for operating the fuel cell.
    Fuel cells are seen as the energy source of the future, especially for mobile use in vehicles of any kind. Here, the proton exchange membrane fuel cell (PEMFC) has emerged as one of the most promising technologies, because it can be operated at low temperatures , offers high response times, has a high power density and can be operated emission-free (reactants only hydrogen and oxygen). However, there are a number of other fuel cell technologies, such as an alkaline fuel cell (AFC), a direct methanol fuel cell (DMFC), a direct ethanol fuel cell (DEFC), a molten carbonate fuel cell (MCFC), a solid oxide fuel cell (SOFC), etc. A fuel cell uses for the anode and for the cathode each a reaction gas, eg Oxygen 02 (or air) and hydrogen H2, which react electrochemically to generate electricity. The structure and functions of the various fuel cells are well known, which is why will not be discussed in detail here. A conditioning of the reaction gases is not mandatory for operation of a fuel cell. However, only with proper gas conditioning can the durability, performance and power density necessary for economical and efficient use of fuel cells, for example in a vehicle, be achieved. Depending on the type of fuel cell, it may be necessary to condition only one of the reaction gases or both reaction gases. The correct process management of a fuel cell, which includes in particular the gas conditioning, is crucial for the performance, durability and safe operation of the fuel cell, especially when exposed to external and internal disturbances. In general, faulty process engineering in fuel cells leads to reversible or irreversible power loss (degradation). An indicator of the current performance of a fuel cell is the state of health (common abbreviation SoH for English State of Health). If a fuel cell reaches a defined value of SoH (in the automobile typically 80% of the continuous power when new) one speaks of end-of-service , which of course is undesirable and should be avoided.
    During gas conditioning, the state variables pressure, temperature and relative humidity (p, T, rH) and the mass flow of the reaction gas are decisive.
    For example, too low a mass flow leads to reactant deficiency, which immediately adversely affects performance and, depending on duration and intensity, causes irreversible damage to the fuel cell. Another important factor is the pressure of the reaction gas. Although a certain pressure gradient between the anode and cathode has a positive influence on the mode of action, however, too large a differential pressure damages the membrane and thus the fuel cell. Another example is the relative humidity of the reaction gas. In a proton exchange membrane fuel cell, for example, it is crucial to protect the membrane from drying out, since only a hydrogenated membrane conducts hydrogen cations and thus is efficient. However, blockade of the gas channels and the diffusion paper due to too much liquid water resulting in reactant undersupply must be avoided at the same time. In addition, a cyclic humidification and dehumidification of the membrane leads to mechanical stress for these and thus again to cracks and defects (pin holes) in the membrane, which favor a direct passage of hydrogen and oxygen. Both effects consequently have a negative influence on the performance and health of a fuel cell. Last but not least, temperature also plays a role. In addition to the accelerated chemical decomposition of the membrane at listening temperatures, the relative humidity and the temperature are also physically coupled, whereby the latter can bring about the above-mentioned effects. The examples given are only an extract of the possible effects of poor gas conditioning and are intended to better understand the problem.
    The major problem for gas conditioning is that the four factors mentioned are dependent on each other due to physical (e.g., thermodynamic) relationships and have non-linear behavior. This problem is often circumvented by the fact that the components of the gas conditioning and the control concept for the gas conditioning are coordinated. As a result, a fairly simple control based on maps, characteristic values, characteristic points, etc. together with simple controllers (such as PID controllers) is largely sufficient. It is also possible that the parameters of the control (maps, characteristic values, characteristic points) are provided with correction factors depending on the bottom.
    If you want to fully exploit the possibilities of a fuel cell, such a simple regulation of gas conditioning is often not sufficient. In particular, as a rule, no (high) dynamic operation of the fuel cell (on the test bench or in the real application) can be realized. Under (high) dynamic operation is understood in particular a rapid response of the scheme, i. that the control is able to follow even rapid changes in the setpoints of the control with the lowest possible control deviation. Especially in the development of a fuel cell on one
    Test bench, where you want to subject the fuel cell usually dynamic test runs (in terms of the rate of change of the parameters, but also the load of the fuel cell) to test or improve the behavior of the fuel cell, this is a problem.
    For a dynamic control of the gas conditioning, this requires a controller which is able to set the controlled variables quickly and accurately and, above all, transiently.
    For this purpose, various approaches to regulating the gas conditioning of a fuel cell can be found in the literature. Many of these approaches are based on a more or less strong simplification of the thermodynamic relationships. Mostly only two of the mentioned influencing variables are regulated and assumptions are made for the other influencing variables. For this purpose, a suitable controller is designed. In most cases, the pressure or humidity is regulated. An example of this is Damour C. et al. "A novel non-linear model-base control strategy to improve PEMFC water management - The flat-ness-based approach", Int. Journal of Hydrogen Energy 40 (2015), p.2371-2376. Therein, a relative humidity controller is designed using the known theory of differential flatness based on a model of membrane moisture. The flatness-based controller shows excellent setpoint tracking behavior, high noise rejection, and high stability. Nevertheless, not all influencing variables can be regulated, which makes this controller unsuitable for the intended regulation of the gas conditioning.
    It is therefore an object of the subject invention to provide a controlled gas conditioning for a reaction gas of a fuel cell, and a corresponding control method for this, which allow an accurate and rapid control of factors influencing the operation of the fuel cell.
    This object is achieved with the features of the independent claims. The invention is based on the fact that the highly nonlinear and coupled multivariable system resulting from the mathematical modeling of the gas conditioning unit can be decoupled and linearized by applying the Lie derivatives. For the resulting linear uncoupled multivariable system, a controller can then be designed using conventional linear control theory. In this way, the gas conditioning unit can be accurately modeled with respect to the influencing variables, which is a prerequisite for an accurate, rapid control of the influencing variables.
    The subject invention will be explained in more detail below with reference to Figures 1 to 4, which show by way of example, schematically and not by way of limitation advantageous embodiments of the invention. It shows
    1 a test stand for a fuel cell with gas conditioning according to the invention, FIG. 2 the course of the output variables with changing input variables of the coupled multivariable system,
    3 shows a controller according to the invention with two degrees of freedom for the gas conditioning and
    4 shows the course of the output variables with changing input variables in accordance with the invention decoupled multi-variable system.
    The invention will be explained below with reference to FIG. 1 without restriction of generality using the example of a test bench 1 for a proton exchange membrane (PEMFC) fuel cell 2. Of course, the fuel cell 2 could also be used as an electrical supply in a machine or plant. Gas conditioning and control would then be implemented in this machine or plant. If, in the following, the operation of a fuel cell 2 is discussed, it is therefore always understood to mean the operation of the fuel cell 2 on a test bench 1 and the real operation of the fuel cell 2 in a machine or installation.
    The fuel cell 2 is constructed in the example of Figure 1 on the test bench 1 and is operated on the test bench 1. As is well known, the fuel cell 2 comprises a cathode C which is supplied with a first reaction gas, for example oxygen, also in the form of air, and an anode A, which is supplied with a second reaction gas, for example hydrogen H2. The two reaction gases are separated from each other inside the fuel cell 2 by a polymer membrane. Between cathode C and anode A, an electrical voltage U can be tapped. This basic structure and function of a fuel cell 2 are well known, which is why will not be discussed further here.
    At least one reaction gas, usually the oxygen-carrying reaction gas, in particular air, is conditioned in a gas conditioning unit 3. In the gas conditioning unit 3, the pressure p, the relative humidity φ, the temperature T and the mass flow m of the conditioned reaction gas are set - in FIG. 1 these four influencing variables are indicated at the inlet of the cathode C. According to the invention, at least three, preferably all four, of these four influencing variables are conditioned. "Conditioning" means that the value of an influencing variable is regulated to a predetermined value, a nominal value. In the case of an influencing variable not conditioned by the gas conditioning unit 3, assumptions can be made for this influencing variable, for example, this influencing variable can be kept constant.
    For controlling these influencing variables, corresponding actuators are provided in the gas conditioning unit 3. In particular, a moistening device 4 for moistening the reaction gas for adjusting a relative humidity φ of the reaction gas, a tempering 5 for temperature control of the reaction gas to adjust a temperature T of the reaction gas, a mass flow controller 6 for controlling the mass flow m of the reaction gas and a pressure control device 7 for regulating the Provided pressure p of the reaction gas. Analogous to the influencing variables to be conditioned, at least three of these four devices, preferably all four, are provided in the gas conditioning unit 3. Of course, a gas source 8 is provided for the reaction gas, which is connected to the gas conditioning unit 3 or is also arranged in the gas conditioning unit 3.
    The gas source 8 is for example a pressure accumulator with compressed, dry reaction gas, for example air. Alternatively, ambient air can also be treated as gas source 8 when using air, for example filtered, compressed, dried, etc.
    The tempering device 5 is, for example, an electrical heating and cooling device or a heat exchanger. As tempering device 5, a device as described in AT 516 385 A1 can also be used.
    The humidifying device 4 in this embodiment comprises a steam generator 9, a mass flow controller 10 for the steam and a mixing chamber 11. As mass flow controller 10 for the water vapor, and also as mass flow controller 6 for the reaction gas, conventional, suitable, commercially available, controllable mass flow controller can be used , In the mixing chamber 11, the water vapor is mixed with the gas originating from the gas source 8 to the conditioned reaction gas for the fuel cell 2.
    Of course, other embodiments of a humidifier 4 are conceivable. For example, water could be supplied to the gas from the gas source 8, e.g. be injected.
    As a pressure control device 7, a back pressure valve is used in this example, which adjusts the pressure p of the reaction gas via the controllable opening cross-section. The back pressure valve 7 is disposed in the gas conditioning unit 3 downstream of the fuel cell 2. This makes it possible to regulate the pressure in front of the fuel cell 2, whereby the
    Pressure control of any pressure losses in the other components of the gas conditioning unit 3 is unaffected.
    After the mixing chamber 11, the reaction gas is in a reaction gas line 12, which is connected to the fuel cell 2, or to the cathode C or anode A of the fuel cell 2, therefore with a certain temperature T, a certain relative humidity φ, a certain pressure p and a certain mass flow m.
    However, this construction of a gas conditioning unit 3 described with reference to FIG. 1 is only an example and of course other constructions of the gas conditioning unit 3 and also other specific embodiments of the humidification device 4, mass flow control device 6, tempering device 5 and pressure control device 7 are conceivable.
    In order to be able to regulate the influencing variables, the moistening device 4, mass flow control device 6, tempering device 5 and pressure regulating device 7 can be controlled via a respective manipulated variable. The manipulated variables are calculated by a control unit 15, in which a controller R is implemented. In the exemplary embodiment shown in FIG. 1, the moistening device 4 is controlled via the mass flow controller 10 for the water vapor with the manipulated variable Us, the mass flow controller 6 with the manipulated variable Ug, the tempering device 5 with the manipulated variable Q and the pressure control device 7 with the manipulated variable uN. With the manipulated variable of the respective actuator is driven to cause a change in the influencing variable. For the controller design of the controller R for controlling the gas conditioning unit 3, a model of the gas conditioning unit 3 is first to be created. Again, various models are conceivable. An advantageous model will be described below, taking into account all four factors. For this purpose, first the system equations for the structure of FIG. 1 are set up.
    From the mass balance in the mixing chamber 11 results
    with the mass mG of the gas, the mass flow of the gas mGin into the mixing chamber 11, the mass flow of the gas mGout from the mixing chamber 11, the mass flow of the water vapor mSin into the mixing chamber 11 and the mass flow of the water vapor mSout from the mixing chamber 11. The mass flow of Gas and water vapor from the mixing chamber 11 is given by
    with the total mass m in the
    Gas conditioning unit 3 and the masses mG of the gas and ms of the water vapor and the mass flow of the reaction gas m. Of course, m = mG + ms.
    From the energy balance of the gas conditioning unit 3 follows
    Here U denotes the internal energy and h the specific enthalpy of the gas (here and below marked by index G), the water vapor (here and in the following marked by index S) and the reaction gas (here and in the following without index) to the mixing chamber 11 and u, denotes the specific internal energy of the gas and the water vapor. The specific enthalpy h of a gas is known to be the product of the specific heat capacity cp at constant pressure and the temperature T of the gas. In the case of steam, the latent heat r0 is additively added. The internal energy u, of a gas is the product of the specific heat capacity cv at constant volumes and the temperature T of the gas. In the case of steam, the latent heat r0 is additively added. If one puts all this into the energy balance and one takes into account the mass balance one obtains the following system equation which describes the temperature dynamics of the gas conditioning unit 3.
    From the thermodynamic equation of state for an ideal gas arises further
    with the pressure p and the temperature T at the entrance of the fuel cell 2. R denotes in a known manner the gas constant for gas (index G), water vapor (index S) or for the
    Reaction gas (without index). The volume V preferably designates not only the volume of the mixing chamber 11, but also the volumes of the piping in the gas conditioning unit 3. The pressure p and the mass flow m of the reaction gas are also significantly influenced by the back pressure valve 7, which can be modeled as follows.
    In this A denotes the opening cross-section of the counterpressure valve 7 and p0 the ambient pressure.
    The relative humidity φ is through
    where Pw (T) designates the saturation partial pressure, for example by
    given is. The parameters pm, Ci, C2 can be, for example, from Plant R.S. et al., "Parameterization of Atmospheric Convection", Vol. 1, Imperial College Press, 2015.
    In addition, the dynamics of the actuators are modeled in dependence on the manipulated variables us, uG, Q and uN in the form of 1st-order delay elements with the time constants τ2 τ3 τ4:
    Therein, TG, o and A0 are predetermined offset quantities.
    From the above system equations one recognizes that a nonlinear multivariable system (ΜΙΜΟ, multiple input multiple output) system of the form
    is present, with the state vector x, input vector u and output vector y as follows:
    In Fig. 1 is indicated for better understanding, where these variables occur in each case.
    The non-linearity results from the system functions f (x), g (x) from the equation of state and the system function h (x) from the output equation, which are each dependent on the state vector x.
    The model of the gas conditioning unit 3 is not only non-linear, but the individual state equations are also coupled several times, whereby the manipulated variables us, uG, Q, uN of the input vector u the outputs T, p, φ, m can not be assigned in the output vector y. This affects several output variables if one of the manipulated variables us, uG, Q, uN changes. This is shown in FIG.
    In FIG. 2, the manipulated variables Us, uG, Q, Un are plotted on the right over the time t and on the left the output variables of the output vector y. At the times 50s, 100s, 150s and 200s, one of the manipulated variables us, uG, Q, uN was changed in each case. In each case, all output variables T, p, φ, m changed. For the coupled, non-linear, ΜΙΜΟ system, a controller must now be designed with which the gas conditioning unit 3 can be regulated. There are many possibilities for this, with a preferred controller design being described below.
    The first step is the nonlinear, coupled multivariable system
    decoupled and linearized. For this, the output, i. an output yj5 in the form
    derived in time, resulting in
    results. Therein, Lfh and Lgh denote the known Lie derivatives of the system function h (x) of the output equation with respect to the system functions f (x) and g (x) of the equation of state of the coupled non-linear multivariable system. The Lie derivatives Lf and Lg are therefore defined as
    It follows from the above that a manipulated variable u has no influence on the respective time derivative of the output variable yj, if Lg hj (x) = 0. Therefore, the output vector y is derived in time for each output variable y as long as the manipulated variable u has an influence on the output variable y, that is to say until the j th derivative. Then, the relative degree of the jth output y is denoted by jj, resulting in the following notation with the Lie derivatives.
    Here Lf denotes the 5th Lie derivative, and (6j) denotes the original temporal derivative.
    If this is applied to all j = 1, ..., m output quantities one obtains in general matrix notation
    with the decoupling matrix J (x). The vector, which contains the zeit time derivatives of the output quantities y, is equated with a new synthetic input vector v, ie
    This results
    It follows that the relationship between the new synthetic input v and the output variables in the output vector y of the multivariable system is decoupled and can be understood as a chain of integrators. If a new synthetic input variable Vj is integrated successive times, the output variable y of the multivariable system is obtained.
    Now a new state vector z is defined in the form
    then one obtains a new multivariable system as a linear, uncoupled state space model according to
    , as well as the (öj xöj) matrix
    The arbitrary linear control theory can then be applied to the linear, uncoupled multivariable system of the form z = Acz + Bcv and any linear regulator can be designed for this purpose. For the control, it is a desired goal that the adjusted output quantities yj follow the predefined setpoint values y d as well as possible (trajectory tracking). The regulation should be as immune to interference as possible. For example, a known two-degree-of-freedom (2DoF) controller R is provided for this purpose, which consists of a feedforward controller FW and a feedback controller FB, for example. is shown in Fig.3. In this case, the feedforward controller FW is to ensure the reference variable behavior (trajectory tracking) and the feedback controller FB to correct any disturbances.
    A new input variable Vj of the decoupled, linear multivariable system 20 corresponds to the öj-th derivative of the output variable y. This results in the feedforward part of the controller R as öj-th derivatives of the setpoints y, dmd. Each setpoint variable y, dmd of the setpoint vector Ydmd is derived according to its relative degree δj and added to the output of the feedback controller FB.
    The feedback controller FB receives in a known manner a control error vector e as a deviation between the setpoint variables y, dmd in the setpoint vector ydmd and the current actual values of the output variables y, ie ej (t) = y (t) -yj, dmd (t). In principle, any feedback controller can be used to correct the error and there are sufficient methods known to determine such a controller. A simple feedback controller FB will be described below.
    The feedback controller FB adjusts to the relative degree öj and the following error parts e x are defined. J'6i
    Thus, a new input variable y of the decoupled, linear multivariable system 20 according to FIG. 3 can be written to as
    For a relative degree öj of two one can produce an analogy to a PID controller, where Kj, o form an integral component, Kj, i a proportional component and Kj 2 a differential component. For a relative degree öj of one would result in a PI controller.
    This results in vectorial notation
    With the above-defined matrices Ac, Bc and
    and
    The controller parameters Kj of the feedback controller FB can then be set to achieve a desired controller behavior. Also for this there are various known approaches, for example the method of Polvorgabe for (Ac - BcK). All desired poles are preferably placed on the left side of the imaginary axis in order to ensure stability. In this way, the controller parameters Kj can be determined. For the above-described linearization and decoupling of the nonlinear, coupled multivariable system, characterized by Σ'1 in FIG. 3, the state variables x are also required, as can be seen from the equations and as also indicated in FIG.
    For this purpose, the state variables x in the gas conditioning unit 3 can be measured, preferably at each sampling instant of the control. Alternatively, the state variables x can also be estimated by an observer from the input quantities u and / or output quantities y, preferably again at each sampling instant of the control. The state variables x can also be calculated in a different, simple way.
    A nonlinear multivariable system of the general form x = f (x, u) has the property of differential flatness when there is a vector of differentially independent outputs y = (yi, ..., ym), so that the state variables x and the inputs u are functions this output y, designated as flat, and its derivatives are:
    It follows that for any arbitrary time curve of the output variable y (t), the associated time profile of the input quantities u (t) and the time profile of the state quantities x (t) can be calculated solely from the time profile of the output variable y (t) without integrating the differential equations of the multivariable system.
    It can be shown that the non-linear multivariable system described above is differentially flat. Thus, the state variables x can be simply calculated from the time profile of the setpoint variables ydmd and do not have to be measured or estimated. This is indicated in Figure 3 by the index F in the state variables x. At the test bench 1, the time profile of the setpoint variables ydmd is determined, for example, by the test run to be carried out and thus known. The state variables xF can thus be calculated in advance offline from the time profile of the setpoint variables ydmd and are then available for the control during the execution of the test run. For the regulation of the gas conditioning unit 3 for the operation of a fuel cell 2, the procedure can be as described below.
    First, if necessary, a regulator R is designed as described above, which has a good Führungsgrößenverhalten, and is stable and robust, so is essentially insensitive to interference. This is achieved, for example, mainly by the choice of the poles of the feedback controller FB. Optionally, however, an already parameterized controller R can also be used. For the operation of the fuel cell 2, a temporal setpoint variation ydmd (t) is specified for the output variables T, p, cp, m serving as set values. This setpoint course ydmd (t) can result from the actual operation of the fuel cell 2, can be predetermined or can be determined by a test run for testing the fuel cell 2 on a test bench. In real operation of the fuel cell 2 in a machine or plant can be calculated, for example, the transition from an operating point to a new operating point, a trajectory along which the fuel cell 2 and the gas conditioning should be performed. This can be done for example in a fuel cell control unit. The fuel cell control unit also provides the operating points from the real operation before. Criteria for the trajectory are e.g. a fast transition, the transient course should not damage the fuel cell 2. The state variables xF (t) can thus be calculated in advance offline from the time profile of the setpoint variables ydmd (t). Alternatively, the state variables x.sub.F (t) can also be calculated or measured online at each sampling instant of the control (ie at each instant at the new manipulated variables). The sampling time for the control is typically in the millisecond range, for example, the control is operated on a test bench 1 with 100 Hz (10 ms sampling time). The gas conditioning unit 3 is then acted upon by the temporal setpoint course ydmd (t), for example according to the test run. For example, desired measurements on the fuel cell 2 can also be carried out on the test stand 1 in order to determine the behavior of the fuel cell 2 during the predetermined test run.
    Such a test run in the form of a predetermined setpoint course ydmd (t) was simulated and the result is shown in FIG. The following parameters were assumed for the simulation: Adjustment range of the humidifier 4 us = [0-30 kgh "1], adjustment range of the tempering device 5 Q = [0-9kW], adjustment range of the mass flow control device 6 Ug = [4-40 kgh" 1] , Adjustment range of pressure regulator 7 uN = [0 - 2 cm2], control range of temperature T = [20 - 100 ° C], control range of pressure p = [1.1 - 3 bar], control range of relative humidity φ = [0 - 100% ] and control range of the mass flow m = [0 - 70 kgh'1]. The following parameters were defined: volume V = 14137 cm3, temperature of water vapor Ts = 141 ° C, ambient pressure p0 = 1 bar, specific heat capacity of gas cp, G = 1-04 kJ kg "1 K'1, specific heat capacity of Water vapor cp, s = 1.89 kJ kg "1 K" 1. The poles of the feedback controller FB were set for the output quantities y with a relative degree öj = 2 to Si = -1, S2,3 = -8 ± 1j and for the output quantities yj with relative degree öj = 1 on Si, 2 = -5.
    In Figure 4 left the predetermined setpoint curve ydmd (t) is shown. To the right of this, the input quantities u are shown, which are set by the controller R. The left-hand diagram also shows the output quantities y calculated in the simulation. On the one hand, one recognizes the excellent reference variable behavior of the controller, that is to say that the output quantities y follow the set values ydmd without recognizable deviations. On the other hand, however, it is also apparent that the individual output quantities y can now be decoupled from one another. Changing one output y leaves the other outputs unaffected. For this purpose, the controller R calculates the combination of the new input variables v, which result in the input variables u, which have to be set with the actuators of the gas conditioning unit 3, at each sampling instant.
    The invention has been explained with reference to the regulation of all four influencing variables or output variables y, ie temperature T, pressure p, relative humidity φ and mass flow m. However, only three of these four parameters can be regulated. Corresponding assumptions can then be made for the fourth, unregulated influencing variable; for example, this fourth influencing variable could be kept constant. With only three controlled influencing variables, the dimension of the above system equations would decrease by one. However, this would not affect the fundamental approach to decoupling and linearization of the coupled, non-linear multivariable system, nor does the described approach of the controller design. However, it would also be possible simply to predefine the desired variable corresponding to the fourth, uncontrolled influencing variable, in accordance with the assumptions. For example, this setpoint variable could be kept constant in the predefined setpoint course ydmd (t).
    The applications of the controlled gas conditioning unit 3 for the gas conditioning of a reaction gas of a fuel cell 2 are manifold. The gas conditioning can in particular both on a test bench (stack or cell test stand), but also in a fuel cell system, for example in a vehicle (ship, train, plane, car, truck, bicycle, motorcycle, etc.), in a power plant (also in force Heat couplings), in emergency power systems, in a handheld device, to any device in which fuel cell systems can be installed. The gas conditioning can thus be used both in real operation of a fuel cell in a fuel cell system, but also on a test bed for testing or developing a fuel cell.
    In principle, gas conditioning can also be used for other applications, for example for conditioning the intake air of an internal combustion engine, again in real operation or on a test bench. But it could also be used to condition gases in process technology, process engineering or medical technology. Similarly, gas conditioning could also be used in metrology to precisely condition a sample gas for accurate measurement.
    权利要求:
    Claims (7)
    [1]
    claims
    1. Regulated gas conditioning for a reaction gas of a fuel cell (2), characterized in that a gas source (8) for a reaction gas of the fuel cell (2), a humidifying device (4) for humidifying the reaction gas for adjusting a relative humidity (cp) of the reaction gas , a tempering device (5) for controlling the temperature of the reaction gas for adjusting a temperature (T) of the reaction gas, a mass flow control device (6) for controlling the mass flow (m) of the reaction gas and a pressure regulating device (7) for regulating the pressure (p) of the reaction gas are that a control unit (15) with a controller (R) is provided, the manipulated variables (uG, us, un, Q) for at least three actuators from the group of humidifying (4), the tempering (5), the mass flow control device ( 6) and the pressure control device (7) calculated so that a predetermined time course of at least three associated influ to set the operating variables of the fuel cell (2) from the group of relative humidity (φ), temperature (T), mass flow (m) and pressure (p) of the reaction gas in a reaction gas line (12). R) on the basis of a model of the gas conditioning unit (3) in the form of a coupled, non-linear multivariable system of the mold

    is designed by designing the controller (R) for the Lie-derivative decoupled and linearized multivariable system.
    [2]
    2. A method for controlling a gas conditioning unit (3) for a fuel cell (2) for operating the fuel cell (2), characterized in that for the control of a controller (R) is used, based on a model of the gas conditioning unit (3) in Form of a coupled, non-linear multivariable system of the mold

    was designed by decoupling and linearizing the coupled non-linear multivariable system using Lie derivatives and designing the controller (R) for the decoupled linear multivariate system, wherein the controller (R) at each sampling instant of the control of predetermined set values (yj, dmd ) the manipulated variables (uG, us, uN, Q) for at least three existing actuators of the gas conditioning unit (3) from the group of a humidifying device (4) for humidifying the reaction gas for adjusting a relative humidity (φ) of the reaction gas, a tempering device (5) for controlling the temperature of the reaction gas for adjusting a temperature (T) of the reaction gas, a mass flow control device (6) for controlling the mass flow (m) of the reaction gas and a pressure control device (7) for regulating the pressure (p) of the reaction gas and calculated that at least these three actuators the gas conditioning unit (3) the calculated control variables (uG, us, un, Q) each set the sampling time of the control.
    [3]
    A method according to claim 2, characterized in that each output (yj) of the coupled non-linear multivariable system is shaped according to its relative degree (δj)

    is derived in time, where the j -th Lie derivative designates that the j-th time derivatives of the output quantities (yj) are equated with new input quantities v, from which a linear, non-coupled state space model of the form z = Acz + Bcv is derived can, with more definite matrices Ac and Bc and a new state vector


    [4]
    Method according to claim 3, characterized in that the controller (R) is designed for the linear, uncoupled state space model of the form z = Acz + Bcv.
    [5]
    5. The method according to claim 4, characterized in that a controller (R) with two degrees of freedom with a feedforward controller (FW) and a feedback controller (FB) is designed, the outputs of the feedforward controller (FW) and the feedback controller ( FB) are added to the new input variables (v) and the feedforward part results from the oj-th time derivatives of the setpoint variables (yj.dmd), and the feedback part the error (e) between the actual values of the output variables (yj) and the values of the setpoint variables (y ^ md).
    [6]
    6. The method according to claim 5, characterized in that error portions (ejg) j j

    be defined, from which

    gives, with the defined matrices Ac and Bc and a matrix K with parameters of the feedback controller (FB), which are set to achieve a desired controller behavior of the feedback controller (FB).
    [7]
    7. The method according to any one of claims 2 to 6, characterized in that the property of the differential flatness of the non-linear, coupled multivariable system, the time course of the state variables (x) according to

    be calculated in advance from the time course of the target values (yj.dmd).
    类似技术:
    公开号 | 公开日 | 专利标题
    EP1267229B1|2007-12-26|Control method and device for startup and shutdown operation of a processing component in a technical process
    AT518518B1|2017-11-15|Controlled gas conditioning for a reaction gas of a fuel cell
    DE102009050938B4|2017-02-16|A method of controlling airflow to a fuel cell stack
    EP3650832A1|2020-05-13|Test chamber for fuel cells and method for controlling
    DE2157722A1|1972-07-20|Electronic proportional controller for regulating the fuel supply in fuel cells
    DE102016106795A1|2016-12-29|The fuel cell system
    DE2509344C3|1980-01-31|Method and arrangement for the automatic control of a boiler-turbine unit
    DE102013001413A1|2013-10-24|Method for controlling inlet temperature of fuel cell in fuel cell vehicle, involves providing modified manipulated variable to moderate temperature system, so as to control inlet temperature of fuel cell
    DE102008055803A1|2009-06-10|A method of model based exhaust mixing control in a fuel cell application
    EP1510758A1|2005-03-02|Method for regulating and/or controlling a burner
    DE102009005647B4|2014-01-09|A fuel cell system and method for controlling the relative humidity of a fuel cell system cathode inlet
    DE3632041A1|1987-04-09|Process and device for regulating the output of a steam power station unit
    EP1205993A1|2002-05-15|Method for operating a fuel cell battery with a control arrangement
    DE102008043869A1|2010-05-20|Control system for fuel cell system, has measurement system for determining state variable of control path, and secondary controller for influencing maximal variable limit and minimal variable limit based on state variable
    DE102006058834A1|2007-08-02|Non-linear cathode inlet / outlet humidity control
    DE3541148C3|1995-12-07|Process for controlling a steam turbine
    EP1454373B8|2009-03-04|Method for operating a pem fuel cell system and corresponding pem fuel cell system
    DE10297104T5|2004-11-04|Method and device for the electrical power control of a fuel cell system
    AT520522B1|2019-05-15|Control of a controlled variable of a conditioning unit of a reactant of a fuel cell with determination of an actual value of the controlled variable
    DE102008010711B4|2018-04-26|Method for operating a fuel cell system and fuel cell system with a regulator assembly
    AT519171B1|2019-08-15|Method and test bench for carrying out a test run for a fuel cell
    DE19931063A1|2001-01-11|Apparatus for cooling and heating a reactor element in an arrangement for producing and/or preparing a fuel for a fuel cell comprises heating/cooling device, conveying device for the heating/cooling medium, and adjusting element
    DE102011080723A1|2013-02-14|Device for supplying reaction media such as cathode air stream to fuel cell array of e.g. motor vehicle, has adjustment device that is provided in each cathode branch duct for regulating volume flow of the supplied reaction medium
    DE3500482C2|1987-07-23|
    DE102007040568A1|2009-03-05|Method for monitoring attitude control of valve slide of constant valve, particular proportional-pressure valve, involves comparing instantaneous value of valve slide with desired value of valve slide in active condition
    同族专利:
    公开号 | 公开日
    CA3031405A1|2018-01-25|
    US10714766B2|2020-07-14|
    WO2018015336A1|2018-01-25|
    AT518518B1|2017-11-15|
    JP2019530129A|2019-10-17|
    EP3488480A1|2019-05-29|
    JP6943943B2|2021-10-06|
    CN109690850A|2019-04-26|
    KR20190030736A|2019-03-22|
    US20190245223A1|2019-08-08|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    WO2006065366A2|2004-12-10|2006-06-22|General Motors Corporation|Nonlinear thermal control of a pem fuel cell stack|
    US20160020476A1|2013-03-11|2016-01-21|University Of Florida Research Foundation, Inc.|Operational control of fuel cells|AT520522A4|2017-12-05|2019-05-15|Avl List Gmbh|Control of a controlled variable of a conditioning unit of a reactant of a fuel cell with determination of an actual value of the controlled variable|JP4618936B2|2001-06-15|2011-01-26|三機工業株式会社|Gas supply device and inspection system|
    AU2003286064A1|2002-11-27|2004-06-18|Hydrogenics Corporation|An electrolyzer module for producing hydrogen for use in a fuel cell power unit|
    DE102004002142A1|2004-01-15|2005-08-11|Robert Bosch Gmbh|Method and computer program for generating a model for the behavior of a control device|
    AU2005251140B2|2004-05-28|2009-12-17|Dcns Sa|Utilization-based fuel cell monitoring and control|
    US7838138B2|2005-09-19|2010-11-23|3M Innovative Properties Company|Fuel cell electrolyte membrane with basic polymer|
    US7829234B2|2005-12-15|2010-11-09|Gm Global Technology Operations, Inc.|Non-linear cathode inlet/outlet humidity control|
    WO2007126308A1|2006-05-01|2007-11-08|Heselmans Johannes Jacobus Mar|Applications for sacrificial anodes|
    US7517600B2|2006-06-01|2009-04-14|Gm Global Technology Operations, Inc.|Multiple pressure regime control to minimize RH excursions during transients|
    CN100545584C|2006-08-10|2009-09-30|上海神力科技有限公司|Be applied to the humidity temperature pickup of fuel cell|
    CN101165506A|2006-10-17|2008-04-23|上海博能同科燃料电池系统有限公司|Fuel battery test system based on network study control|
    US8173311B2|2007-02-26|2012-05-08|GM Global Technology Operations LLC|Method for dynamic adaptive relative humidity control in the cathode of a fuel cell stack|
    DE102007014616A1|2007-03-23|2008-09-25|Forschungszentrum Jülich GmbH|Fuel cell system and method for controlling a fuel cell system|
    DE102008024513B4|2008-05-21|2017-08-24|Liebherr-Werk Nenzing Gmbh|Crane control with active coast sequence|
    CN102754264A|2011-01-28|2012-10-24|丰田自动车株式会社|Fuel cell system|
    DE102012218636A1|2012-10-12|2014-04-17|Robert Bosch Gmbh|Determination of the fuel cell input humidity via pressure sensors and a mass flow-dependent control of the humidifier bypass|
    CN102968056A|2012-12-07|2013-03-13|上海电机学院|Modeling system of proton exchange membrane fuel cell and intelligent predictive control method thereof|
    CN105653827B|2016-03-17|2020-03-13|北京工业大学|Hypersonic aircraft Terminal sliding mode controller design method|CN113140765A|2021-03-04|2021-07-20|同济大学|Fuel cell air inlet flow and pressure decoupling control method and system|
    法律状态:
    优先权:
    申请号 | 申请日 | 专利标题
    ATA50663/2016A|AT518518B1|2016-07-20|2016-07-20|Controlled gas conditioning for a reaction gas of a fuel cell|ATA50663/2016A| AT518518B1|2016-07-20|2016-07-20|Controlled gas conditioning for a reaction gas of a fuel cell|
    KR1020197004915A| KR20190030736A|2016-07-20|2017-07-17|Gas conditioning controlled for the reactive gas of a fuel cell|
    EP17743004.8A| EP3488480A1|2016-07-20|2017-07-17|Regulated gas conditioning process for a reaction gas of a fuel cell|
    PCT/EP2017/067999| WO2018015336A1|2016-07-20|2017-07-17|Regulated gas conditioning process for a reaction gas of a fuel cell|
    CA3031405A| CA3031405A1|2016-07-20|2017-07-17|Controlled gas conditioning for a reaction gas of a fuel cell|
    CN201780044244.1A| CN109690850A|2016-07-20|2017-07-17|The controlled air regulating device of reaction gas for fuel cell|
    JP2019502570A| JP6943943B2|2016-07-20|2017-07-17|Controlled gas regulator for fuel cell reaction gas|
    US16/318,338| US10714766B2|2016-07-20|2017-07-17|Controlled gas conditioning for a reaction gas of a fuel cell|
    [返回顶部]