DE102016206512A1 - Vehicle and method for operating a vehicle - Google Patents

Abstract

The invention relates to a vehicle (200) having a fuel cell system (100), comprising a fuel cell stack (10) for providing an electrical power Pstack, at least one auxiliary unit (24, 26, 33, 34, 38) with electrical power consumption Paux for operating the fuel cell stack (10) and a net performance Pnetto = Pstack - Paux. The vehicle also has at least one consumer (44, 51) with an electrical power request Puse dependent on a control value x (t) and a control unit (60). The control unit (60) is set up to determine a course of the detected manipulated variable and a predicted manipulated variable x * (tzi) from values of the manipulated variable x (tvi) acquired over a specific period, a predicted power request from the predicted manipulated variable x * (tzi) and thus to determine a predicted stack performance. The control unit is further configured to increase a current stacking power Pstack (t) by - Pstack (t).

Description

The invention relates to a vehicle, comprising a fuel cell system, and a method for operating such a vehicle.

Fuel cells use the chemical transformation of a fuel with oxygen to water to generate electrical energy. For this purpose, fuel cells contain as a core component the so-called membrane electrode assembly (MEA for membrane electrode assembly), which is a microstructure of an ion-conducting (usually proton-conducting) membrane and in each case on both sides of the membrane arranged catalytic electrode (anode and cathode).

As a rule, the fuel cell is formed by a multiplicity of stacked MEAs whose electrical powers add up. As a rule, bipolar plates (also called flow field plates or separator plates) are arranged between the individual membrane electrode assemblies, which ensure that the individual cells are supplied with the operating media, ie the reactants, and are usually also used for cooling. In addition, the bipolar plates provide an electrically conductive contact to the membrane-electrode assemblies.

The output from a fuel cell _stack electric power P _stack usually depends on a given to the fuel cell power request P _use , that is, the current depth of a downstream current sink, and the fuel cell stack supplied resource streams, in particular the cathode and anode operating current from. When using a fuel cell stack for providing an electrical power is also to be considered that for the operation of a fuel cell stack, a plurality of ancillaries is necessary, which also consume an electric power P _aux .

Depending on the fuel cell system, the plurality of ancillaries may include at least one of an air compressor, recirculation fan, cooling water pump, valves, sensors, etc. The power consumption of these components can be referred to as parasitic power consumption P _aux , since it must be provided by the fuel cell stack, but is not available for external consumers. The net power of the fuel cell system P _net available for external consumers thus results as a difference between the electric power P _stack generated by the _stack and the parasitic power consumption P _aux . The standing for external loads (such as for the electric drive) available net power according to P = P _{net stack} - P _aux is thus always below the performance of the fuel cell stack P _stack.

A vehicle equipped with a fuel cell system generally has a control unit for operating the fuel cell system, in particular its anode and cathode supply. The control unit detects as a control value usually requested by the driver of the vehicle driving power P _W , for example, the strength of an accelerator pedal operation. This requested driving performance can also be described as a desired performance. In addition, it can capture power requests of other electrical consumers of the vehicle, such as an air conditioner. In response to these input variables, in particular the requested driving power P _W , the control unit immediately determines a power P _{stack to be provided} by the fuel cell _stack or, initially, a net power P _{net to be provided} .

If a power P _{stack to} be provided is determined directly from the power request of the at least one consumer, a net power P _{net is} actually output from the fuel cell _stack which is reduced by the power consumption of at least one auxiliary unit P _aux (P _net = P _{stack -P aux} ). By determining the net power output from the stack, for example by measuring the current delivered by the stack, the power requested by the fuel cell stack can be readjusted until the delivered net power meets the power request. Alternatively, the control unit determines a power consumption P _{aux of} the at least one ancillary unit necessary for providing a stacking capacity in the amount of the power request, for example by means of a characteristic map (LUT), and finally the total power P _{stack to be} requested by the fuel cell _stack according to P _stack = P _net + P _aux . Based on the stacking power P _{stack to} be provided, the control unit calculates or stores the required mass flows and / or operating pressures of the anode and cathode operating medium and controls or regulates the ancillaries of the fuel cell system accordingly, in particular a compressor for supplying the cathode side with air and adjusting means for supply the anode side with hydrogen. For the purposes of this application, the control unit is as a functional unit too understand, with the implementation of the functions in components for vehicle control and / or in components of the fuel cell system can take place.

Known fuel cell systems typically have some inertia in dealing with large power requests, such as sudden acceleration of a fuel cell powered vehicle. The inertia results, on the one hand, from the inertia of the operating mass flows to be moved and from the inertia of the ancillary components themselves. For example, the cathode-side compressor must be greatly accelerated in order to convey the required increased cathode operating medium flow. Due to this acceleration, the electrical power consumption of the at least one secondary consumer P _{aux increases} up to a local maximum. As a result, the net power P _{net of} the fuel cell _{stack has} a local minimum compared to the _stack power P _stack .

The fuel cell system may also have some inertia in generating a power request P _{use of} an associated consumer. The power request P _use is usually determined by means of a transfer function from a time-dependent control value. For example, if a driver requests a maximum acceleration of a stationary vehicle by means of a pedal value, the pedal value initially acts as a control value for a torque request M _use . The power request P _{use of} an electric motor is determined according to P _use = 2 · π · M _use · n · η and thus also depends on a current engine speed. Thus, the power request of the electric motor increases with increase in the engine speed and is thus delayed from the manipulated value or the torque request.

It is an object of the present invention to overcome the shortcomings of the prior art and to propose a fuel cell system powered vehicle and method of operation, wherein inertia reduces the increase in fuel cell _stack power P _{stack in} response to an increased power request is.

The object is achieved by a vehicle and a method for operating such a vehicle having the features of the independent claims.

The object according to the invention is achieved by a vehicle having a fuel cell system, comprising a fuel cell stack for providing an electrical power P _stack , at least one accessory with electrical power consumption P _aux for operating the fuel cell stack and a net power P _net = P _{stack -P aux} . The vehicle also has at least one consumer and a control unit. During operation of the vehicle, the at least one consumer makes an electrical power request to the fuel cell stack, which depends on a manipulated variable x (t).

According to the invention, the control unit is set up to detect values of the control value over a certain period of time or to read in otherwise detected control values. The control unit is further configured to determine a course of the control value and to use the current value x (t) of the control value and the determined curve of the control value for a future time t _zl to determine a predicted control _value x * (t _zl ). The control unit is further configured to use the predicted control value to determine a predicted power request P * _{use of} the at least one consumer.

Without being limited thereto, the course of the manipulated variable is preferably the gradient x (t) of the manipulated variable. The control unit is likewise preferred for determining from the detected values of the manipulated variable a course of the manipulated variable by means of regression analysis, for example by means of polynomial functions of the nth order, by means of a moving average and / or by comparison with stored progression patterns. For reasons of clarity, the course of the manipulated variable based on the gradient of the detected manipulated variable is explained below, but without being limited thereto. The control unit is preferably set up on the basis of the current value x (t) of the control value and of the determined gradient x. (T) of the control value for a future time t _zl to determine a predicted control _value x * (t _zl ).

According to the invention, the control unit is set up to determine a predicted stack power P * _stack on the basis of P * _use and to increase the current _stack power P _stack (t) ^by ΔPstack = P * _stack -P _stack (t) to the predicted stack power P * _stack , The control unit is preferably set up to determine a predicted stacking power P * _stack = P * _aux + P * _net . The predicted net power P * _net corresponds to the predicted power _request P * _use and the predicted power consumption of the at least one auxiliary unit P * _{aux is} determined as a function of the predicted net power P * _net , for example from a performance map. The control unit is then configured to increase the current _stack power P _stack by ΔP _stack = P * _stack -P _stack (t) to the predicted stack power P * _stack = P * _net + P * _aux .

During operation of the fuel cell system of the vehicle, a current power request of the at least one load is determined as a function of a current control value by means of at least one transmission function. For example, from a pedal value first a torque request and therefrom an electric power request of an electric motor can be determined. Subsequently, a stacking power is determined which is necessary to fulfill this power request. If this stacking power is requested by the fuel cell stack, it preferably provides a net power corresponding to the power request. The control unit according to the invention is set up for this determination of the electrical power request from the control value. Particularly preferably, the transfer function (s) used in this context are also used to determine the predicted service _request P * _use from the predicted manipulated _variable x * (t _zl ). By requesting an increased stacking capacity at an early stage by means of the predicted power request, this is advantageously already available when increasing the current control value or the current power request.

In a vehicle according to the invention, the inertia is minimized when increasing the power provided to a consumer of the vehicle, for example the electric motor, by the fuel cell system. The prediction of the manipulated variable, which is at the beginning of a signal chain extending from the driver's request to the wheel moment, starts directly at the driver's request. Thus, compared to forecasts based on past power requests or stacking performances, system inherent inertia in creating the power request or providing the stack power is not included in the stack power forecast. The fuel cell system provides the vehicle of the invention with a stacking power based on a predicted power request based on a predicted manipulated variable before the power request is actually polled by adjusting the setpoint. At the time of changing the control value and increasing the power request, the fuel cell stack thus advantageously already provides the target power.

In a preferred embodiment of the vehicle according to the invention, the control unit is further configured, before increasing the stacking capacity by ΔP _stack, a predicted maximum power consumption P _aux ^{max of} the at least one accessory when increasing the stacking capacity by ΔP _stack and, with P _aux ^max + P _net ( t) = P _net ⁱⁿ and P _aux ^im = f (P _net ⁱⁿ ), to determine a transition _stack performance P _stack ⁱⁿ = P _net ⁱⁿ + P _aux ^im and the current _stack power P _stack (t) first to P _stack ⁱⁿ and then around ΔP _stack ⁱⁿ = P * _stack - P _stack ⁱⁿ increase.

The control unit of the vehicle according to the invention is thus preferably adapted before raising the stack power to .DELTA.P _stack a required for this predicted maximum power P ^max of the _aux to determine at least one secondary assembly. In other words, the control unit calculates a maximum of the power consumption of the at least one auxiliary unit when increasing the stacking capacity by ΔP _stack , that is from the current to the predicted stacking capacity, or reads it from a characteristic field. The actual stacking capacity is initially not increased. The control unit then sets the sum of the predicted maximum power P _aux ^max and a current stack net power P _net (t) as a predicted transition stack net power P _net ⁱⁿ and determines a transition stack power P _stack ⁱⁿ = P _net ^{in the} + P _aux ⁱⁿ by a transitional power consumption of the at least one accessory P _aux ^im as a function of P _net ⁱⁿ calculated or read from a map.

According to this embodiment, the control unit is further configured to first increase the current _stack power P _stack (t) to the transition _{stack power} P _stack ^im and then to the predicted stack power P * _stack . Preferably, the controller increases the _{stack power} to P * _stack at the earliest when responsive to the request for _stack power P _stack ⁱⁿ a transition _{stack net} power P _net ⁱⁿ which is the sum of the projected maximum power consumption P _aux ^max and a previously required stacking power P _net (t <t _im ) actually corresponds. A local minimum of the stack net power in increasing the stack power of P ^{in the} _stack on the predicted stack power P * _stack is advantageously avoided according to this embodiment. Advantageously, inertia associated with starting up at least one ancillary unit when increasing a requested stacking capacity is significantly reduced.

According to a particularly preferred embodiment, the control unit is configured to recursively perform the prediction of P _aux ^max and P _stack ⁱⁿ and the step increase of the stack performance. That is, the controller determines at least a second predicted maximum power consumption P * _aux ^{max of} the at least one accessory when increasing the current stack power to P _stack ⁱⁿ and deriving therefrom at least a second predicted stack transition power P * _stack ^im , as described above. Subsequently, the control unit increases the current stacking power first to P * _stack ⁱⁿ , then to P ⁱⁿ and finally to _stack P * _stack . Thus, a significantly linearized and accelerated increase in stacking horsepower is achieved as a result of the predicted power request P * _use .

Also preferably, the manipulated variable x (t) is a driver-specific variable, that is to say a quantity that reflects a driver's request regarding a power request to the at least one consumer. The manipulated variable x (t) is particularly preferably a pedal angle or a pedal angle gradient which is detected, for example, by means of a pedal value transmitter and read in by the control unit. Likewise, the at least one manipulated variable may also be a variable output by a cruise control or another device for speed regulation of a vehicle, for example as a function of distances to other vehicles or road markings.

In a preferred embodiment of the vehicle according to the invention, the control unit is further configured to determine the predicted power _request P * _use as a function of the predicted control _value x * (t _zl ) and as a function of at least one vehicle _parameter . The vehicle parameter may be a current and / or a predicted vehicle parameter. A predicted manipulated _variable x * (t _zl ) can thus lead to different predicted power _requests P * _use (t _zl ) depending on the vehicle state. Thus, the prediction accuracy can be improved with respect to a power _request actually made at the future time t _zl .

Particularly preferably, the at least one consumer of the fuel cell system has an electric motor and the current vehicle parameter is its rotational speed n. The power request of an electric motor is a function of its current rotational speed and its current torque. Thus, by using the current speed of an electric motor, a predicted control value, for example a predicted pedal position, can be designed differently, that is, lead to different predicted power requests. The current vehicle parameter is also preferably a wheel speed, vehicle speed and / or vehicle acceleration. These parameters are readily convertible into the speed or the speed gradient of the electric motor and allow the conclusions already described.

In a likewise preferred embodiment, the control unit is further configured to incorporate a predicted vehicle _parameter , for example n * (t _zl ), into the determination of a predicted power _request P * use (t _zl ) by means of a predicted control _value x * (t _zl ). A predicted vehicle parameter, for example a predicted engine torque, is preferably determined on the basis of a profile, particularly preferably of a gradient, of detected values, for example on the basis of detected torques or detected rotational speeds and power values. For example, a predicted setpoint results in a greater increase in the predicted power request if the recently detected torques were sharply increasing and not constant.

The prediction of the at least one vehicle parameter likewise preferably takes place on the basis of current and / or predicted environmental data. Current environmental data is preferably detected by means of at least one sensor, for example a camera or a gyroscope. Predicted environmental data are preferably determined from navigation data, in particular an expected route profile.

Particularly preferably, the environmental data is a road gradient or a curve radius. On a downgrade, a specific wheel and engine speed is reached earlier than on a level track. A predicted power request P * _use must therefore rise more quickly in response to a specific predicted manipulated variable x * (t) than at a level distance. At the same time, based on the predicted control value, for example a predicted pedal angle, a lower predicted power request can be determined, since the gradient contributes to the acceleration. With an increase, the engine speed and engine power are delayed with respect to the plane. Thus, a predicted power request P * _use must rise more slowly in response to a certain predicted manipulated variable x * (t) than at the level. At the same time, based on the predicted control value, for example a predicted pedal angle, an increased predicted power request can be determined, since the increase counteracts the acceleration. Curve radii are determined by limit speeds that are not to be exceeded and the predicted power request is limited. This embodiment increases energy efficiency, for example, when going uphill. At the same time, an inertia of the power supply is reduced, for example when the gradient is expected.

According to a preferred embodiment of the vehicle according to the invention, the control unit is further configured to generate at least one and preferably a plurality of time-dependent control values x _j (ti) over a specific period Δt = {t _vi , t _{vi + 1} ,..., T _{v ( i + k)} , ..., t}, in particular to record a multiplicity of past times t or to read set values registered at these times. The control unit is further configured to at least one or a plurality of courses of the detected control values, particularly preferably gradients from the detected control values according to x. _j (t, t _vi ) = (x _j (t) -x _j (t _vi )) / (t -t _vi ) or according to x. _j (t _{v (i + k)} , t _vi ) = (x _j (t _{v (i + k)} ) - x _j (t _vi )) / (t _{v (i + k)} - t _vi ) , In other words, the control unit can determine the at least one course, particularly preferably the at least one gradient, from a dynamic that continues to the current time and / or from a previously learned dynamic. The choice of the time segments for determining the course, particularly preferably the gradient, can be made dependent on the current operating time of the fuel cell stack and / or the vehicle and / or environmental data, for example.

The control unit is further set up for at least one and preferably a plurality of future times t _zl at least one predicted set _value according to x _j * (t, t _zl ) = x _j (t) + x. _j (t _zl - t _vi ) * (t _zl - t) or according to x _j * (t, t _zl ) = x. _j (t _{v (i + k)} (t _vi) * (t _zl - t) to determine the control unit is thus preferably adapted to the dynamics of at least one or a plurality of control values during different periods of time (t -. t _vi) or (t _{v (i + k)} - t _vi).. mapped by means of gradients The control unit is preferably further configured with this gradient x _j (t _{v (i + k), vi} t) x = (i, j. k) or _x.j (t, t _vi ), t _vi ) = x. (i, j) a plurality of predicted manipulated _values x i (i, j, l), in particular for different future times t _zl for the at least one or to determine each of the plurality of manipulated values x _j . Thus, captured data are used optimally and the stability of the prognosis is increased.

Particularly preferably, the control unit is set up for determining one or more gradients x. _j (t, t _vi ) or x. _j (t _{v (i + k)} , t _vi ) to vary the at least one past time t _vi or t _{v (i + k)} or, in other words, the size of a past time window under consideration. The control unit is likewise preferably configured to vary the at least one future time t _z1 or, in other words, the size of the considered future time window when determining one or more predicted manipulated _values x * (i, j, l). Thus, the control unit is set up to selectively consider and / or influence the short-term behavior and / or the long-term behavior of a vehicle dynamics. In addition, the considered time windows can be influenced as a function of the sign of a gradient. For example, the time window t _zl can be _selected to be large in the case of accelerations and small in the case of delays. For this, the parallel consideration of multiple time windows is necessary.

Preferably, the past time t _vi or the past times t _{v (i + k)} and t _vi and the future time t _{zl are selected} as a function of, for example set by the driver, operating mode. In a sportive mode of operation, the short-term behavior of the vehicle can be exploited for the far-reaching prediction of increased power to be provided, for example by selecting (t-t _vi ) small and (t _zl -t) large. The focus is on the future availability of high performance through long-term predictions of increased power _{request usage} based on a recent change in setpoint, such as recent acceleration. In an ecological mode of operation, the long-term behavior of the vehicle can be used for the short-term provision of a previously requested power by selecting (t-t _vi ) large and (t _zl -t) small. The focus here is on avoiding a misprediction of an increased _{performance, which is} actually not requested at time t _zl .

Also preferred are the past time t _vi or the past times t _{v (i + k)} and t _vi and / or the future time t _zl depending on an operating time, ie a current operating time and / or a total operating time, the fuel cell stack and / or the vehicle selected. Thus, a changed dynamics in the power delivery of a fuel cell stack can be compensated. For example, the time window (t _zl -t) can be successively increased as the total _operating time of the stack increases, in order to successively increase the time span bridged by the precontrol and to provide the stack with sufficient time for the high-edge of a power request over its entire service life.

In a particularly preferred embodiment of the vehicle according to the invention, the control unit is further adapted to at least one predicted power request depending on the at least one and preferably on the plurality of predicted control values x * _j (t, t _li ) = x (i, j, l) P * _use (x * _j (t, t _zl )) = P * _use (i, j, l). The at least one and preferably the plurality of predicted setting values preferably comprises predicted values of different setting values, which are particularly preferably in each case based on different past time windows (t-t _vi ) or (t _{v (i + k)} -t _vi ) and for different future times t _{zl are} determined.

On the basis of the at least one or a plurality of predicted power requests P * _use (i, j, l), the control unit determines at least one or a plurality of predicted stacking powers P * _stack (i, j, l) = P * _net (i, j, l) + P * _aux (i, j, l). The predicted stacking power P * _net (i, j, l) is set equal to P * _use (i, j, l) and the predicted power consumption P * _net (i, j, l) is dependent on P * _use (i, j, l) or read from a map. The control unit is finally configured to change the current stacking power P _stack (t) by ΔP _stack (t, t _zl ) = (1 / Σ _{i, j, l} a ^j , _{i, l} ) Σ _{i, j, l} (P * _stack (i, j, l) - P _stack (t)) increase. A pre-control value for increasing the current stack power is thus calculated as a weighted average of the plurality of predicted stack power P * _stack (i, j, l). Different past time intervals t _vi , different future time intervals t _zl and / or different position values x _{j can} enter into the prognosis differently by means of the weighting factors a ^j , _{i, l} . The weighting factors are preferably time-dependent. Without weighting factors, ΔP _stack (t, t _zl ) = 1 / (n _i + n _j + n _l ) Σ _{i, j, l} ^aj , _{i, l} (P * _stack (i, j, l) -P _stack (t)) Long-term behavior and short-term behavior of the vehicle are thus equally in the pre-control of the power request of at least one consumer, which increases the quality and stability of the feedforward.

The invention likewise relates to a method for operating a vehicle, wherein the vehicle is a fuel cell system with a fuel cell stack for providing an electric power P _stack , at least one accessory with electrical power P _aux for operating the fuel cell stack and a net power P _net = P _stack - P _aux . The vehicle also has at least one consumer with an electric power request P _use dependent on a control value x (t) and a control unit. The method according to the invention comprises the following method steps: detection or reading in of the manipulated variable x (t _vi ) over a specific period Δt = {t _vi , t _{vi + 1} ,..., T _{v (i + k)} ,. t}, determining a course, preferably a gradient x. _j (t, t _vi ) or x. _j (t _{v (i + k)} , t _vi ), the detected control _value and thus determining at least one predicted control _value x * (t _zl ).

The method further comprises the steps of determining at least one predicted power _request P * _use as a function of the at least one predicted control _value x * (t _zl ) and determining a predicted stack power P * _stack as a function of the at least one predicted power request P * _use . Preferably, a predicted stacking power P * _stack = P * _net + P * _{aux is} determined by setting the predicted stacking capacity P * _net equal to P * _use and determining the predicted power consumption of the ancillary components P * _aux as a function of the stacking _{net power} P * _net according to P * _aux = f (P * _net ). The inventive method further comprises increasing a current _stack power P _stack (t) by ΔP _stack = P * _stack -P _stack (t). Preferably, the inventive method comprises the steps of: determining a predicted maximum power P _aux ^max of the at least one secondary consumer in increasing the stack power to .DELTA.P _stack, determining a transition stack power P _stack ⁱⁿ = P _net ^{in the} + P _aux ⁱⁿ by setting a transitional stack net power P _net ⁱⁿ equals the sum of the predicted maximum power consumption P _aux ^max and a current stacking _net power P _net (t) determining the predicted transient power consumption of the ancillary units P _aux ^im as a function of the transition stack _{net power} P _net ⁱⁿ P _aux ^im = f (P _net ^im ) and finally increasing one current stacking power P _stack (t) first at P _stack ⁱⁿ and then at ΔP _stack ^at (t) = P * _stack - P _stack ^at .

In addition to the power of an electric motor, the desired charging power of an energy store or the power request of secondary consumers can also be added to P _use .

Likewise provided by the invention is a vehicle comprising a fuel cell system with a fuel cell stack for providing an electric power P _stack , at least one accessory with electrical power consumption P _aux for operating the fuel cell stack and a net power P _net = P _stack - P _aux , at least one consumer an electric power request P _use dependent on a control value x (t) and a control unit, wherein the control unit is adapted to increase a predicted maximum power consumption P _aux ^{max of} the at least one auxiliary unit before increasing a current _stack power P _stack (t) by ΔP _stack Increasing the stacking power by ΔP _stack and, with P _aux ^max + P _net (t) = P _net ⁱⁿ and P _aux ^im = f (P _net ⁱⁿ ) to determine a transition _{stack power} P _stack ⁱⁿ = P _net ⁱⁿ + P _aux ⁱⁿ and the current stacking power P _stack (t) first to P _stack ⁱⁿ and then to ΔP _stack ⁱⁿ = ΔP _stack + P _stack (t) - P _stack ⁱⁿ to increase.

In preferred embodiments, the fuel cell system further comprises at least one consumer with an electric power request P _use dependent on a control value x (t), and the control unit of this vehicle is preferably set up as described above. In particular, the control unit of this vehicle is preferably also configured to determine a course of the detected manipulated _variable and thus a predicted manipulated _variable x * (t _zl ) from values of the _{manipulated variable} x (t _vi ) recorded over a specific period of time, from the predicted manipulated _variable x * (t _zl ) a predicted power _request P * _use and from this to determine a predicted stacking power P * _stack and to increase a current stacking power P _stack (t) by ΔP _stack = P * _stack -P _stack (t).

Further preferred embodiments of the invention will become apparent from the remaining, mentioned in the dependent claims characteristics.

The various embodiments of the invention mentioned in this application are, unless otherwise stated in the individual case, advantageously combinable with each other.

The invention will be explained below in embodiments with reference to the accompanying drawings. Show it:

1 a schematic representation of a vehicle according to an embodiment;

2 a schematic representation of a vehicle according to an embodiment; and

3 a control value course, the course of a corresponding power request and a predicted control value;

4 a control value curve and an alternatively predicted control value;

5 a manipulated variable course, the course of a corresponding power request and the course of a predicted power request according to an embodiment of the invention; and

6 the course of a power request, a predicted power request according to an embodiment of the invention, a stack performance, a stacking net power and a power consumption of at least one accessory when increasing the power request.

1 shows a total of 200 designated vehicle according to a preferred embodiment of the present invention. The emphasis of the presentation lies on the supply side on the total with 100 designated fuel cell system as part of the vehicle 200 , The vehicle is in particular an electric vehicle with one of the fuel cell system 100 electric motor supplied with electric power.

The fuel cell system 100 comprises as a core component a fuel cell stack 10 , with a plurality of stacked individual cells 11 Interrupted by Alternately Stacked Membrane Electrode Arrangements (MEA) 14 and bipolar plates 15 are formed (see detail). Every single cell 11 each includes an MEA 14 with an ion-conducting polymer electrolyte membrane (not shown here) and catalytic electrodes arranged on both sides. The electrodes catalyze the respective partial reaction of the fuel conversion. The anode and cathode electrodes are formed as a coating on the membrane and comprise a catalytic material, for example platinum, supported on an electrically conductive substrate of high specific surface area, for example a carbon based material.

As in the detailed representation of the 1 is shown between a bipolar plate 15 and the anode an anode compartment 12 is formed and is between the cathode and the next bipolar plate 15 the cathode compartment 13 educated. The bipolar plates 15 serve to supply the resources in the anode and cathode spaces 12 . 13 and further provide the electrical connection between the individual fuel cells 11 ago. Optionally, there are gas diffusion layers between the membrane-electrode assembly 14 and the bipolar plates 15 arranged.

The fuel cell system 100 has to supply the fuel cell stack 10 with the resources an anode supply 20 and a cathode supply 30 on.

The anode supply 20 of in 1 shown fuel cell system 100 includes an anode supply path 21 for supplying an anode operating agent (the fuel), for example hydrogen, into the anode spaces 12 of the fuel cell stack 10 , The anode supply paths 21 connect a fuel storage 23 with an anode inlet of the fuel cell stack 10 , The anode supply 20 further includes an anode exhaust path 22 , for discharging the anode exhaust gas from the anode chambers 12 and via an anode outlet of the fuel cell stack 10 , The anode operating pressure on the anode sides 12 of the fuel cell stack 10 is about a first actuating means 24 in the anode supply path 21 adjustable.

The anode supply 20 of in 1 shown fuel cell system also has a recirculation line 25 on, the the anode exhaust path 22 with the anode supply path 21 combines. The recirculation of fuel is used to fuel the stack of fuel stoichiometrically 10 due. In the recirculation line 25 is a recirculation conveyor 26 , preferably a recirculation blower, arranged.

The cathode supply 30 of the fuel cell system 100 of the 1 further includes a cathode supply path 31 for supplying an oxygen-containing cathode resource to the cathode spaces 13 of the fuel cell stack 10 , This is in particular air, which is sucked from the environment. The cathode supply 30 further includes a cathode exhaust path 32 , which the cathode exhaust gas (in particular the exhaust air) from the cathode compartments 13 of the fuel cell stack 10 dissipates and optionally this feeds an exhaust system, not shown. In the cathode supply path 31 is also a compressor 33 arranged to convey and compress the cathode operating means. In the illustrated embodiment, the compressor 33 about one with a corresponding power electronics 35 equipped electric motor 34 driven. The compressor 33 may also be through a in the cathode exhaust path 32 arranged turbine 36 supportively driven by a common shaft (not shown).

This in 1 shown fuel cell system 100 also has a humidifier module 39 on. The humidifier module 39 on the one hand is in the cathode supply path 31 arranged so that it can be flowed through by the cathode operating gas. On the other hand, it is so in the cathode exhaust path 32 arranged so that it can be flowed through by the cathode exhaust gas. A humidifier 39 typically has a plurality of water vapor permeable membranes formed either flat or in the form of hollow fibers. In this case, one side of the membranes is overflowed by the comparatively dry cathode operating gas (air) and the other side by the comparatively moist cathode exhaust gas (exhaust gas). Driven by the higher partial pressure of water vapor in the cathode exhaust gas, there is a transfer of water vapor across the membrane in the cathode operating gas, which is moistened in this way. The cathode supply 30 also has a bypass line 37 on which the cathode supply line 31 with the cathode exhaust gas line 32 combines. One in the bypass line 37 arranged adjusting means 38 serves to control the amount of the fuel cell stack 10 immediate cathode resource.

Various other details of the anode and cathode supply 20 . 30 are in the simplified 1 not shown for reasons of clarity. Thus, in the anode and / or cathode exhaust path 22 . 32 a water separator may be installed to condense and drain the product water resulting from the fuel cell reaction. Finally, the anode exhaust gas line 22 into the cathode exhaust gas line 32 lead, so that the anode exhaust gas and the cathode exhaust gas are discharged via a common exhaust system.

The vehicle 200 also has at least one consumer 44 . 51 with electrical power request P _use and a control unit 60 for providing a power request to the fuel cell stack 10 on. The control unit 60 puts the power request to the fuel cell stack 10 preferably at least depending on the power request.

2 also shows a schematic representation of the vehicle 200 according to a preferred embodiment of the present invention. The focus of the presentation lies on the consumer side on the function of the control unit 60 in connection with the at least one consumer 44 . 51 ,

This in 2 shown vehicle 200 has the reference to 1 detailed fuel cell system 100 , the electronic control unit 60 , an electrical power system 40 and a vehicle drive system 50 on.

The electric power system 40 includes a voltage sensor 41 for detecting one of the fuel cell stack 10 generated voltage and a current sensor 42 for detecting one of the fuel cell stacks 10 generated electricity. Further, the electric power system includes 40 an energy store 44 , For example, a high-voltage battery or a capacitor. In the performance system 40 is also a converter implemented in triport topology 45 (Triport converter) arranged. To a first side of the dual DC / DC converter 45 is the battery 44 connected. All traction network components of the drive system 50 are at fixed voltage on a second side of the converter 45 connected. In the same or similar manner, the ancillaries of the fuel cell system itself, for example, the electric motor 34 of the compressor 33 (please refer 1 ), or other electrical consumers of the vehicle, such as a compressor for an air conditioner and the like, to be connected to the power grid.

The drive system 50 includes an electric motor 51 acting as a traction engine of the vehicle 200 serves. For this drives the electric motor 51 a drive axle 52 with drive wheels arranged thereon 53 at. The traction engine 51 is about an inverter 43 with the electronic power system 40 of the fuel cell system 100 connected and represents the main electrical consumer of the system.

The electronic control unit 60 controls the operation of the vehicle 200 and in particular the fuel cell system 100 including its anode and cathode supplies 20 . 30 , the electric power system 40 and the traction engine 51 , For this purpose, the control unit receives 60 various input signals, such as those with the voltage sensor 41 detected voltage U of the fuel cell 10 with the current sensor 42 detected current I of the fuel cell stack 10 (from which the control unit calculates the current stack voltage P _stack (t)), information about the temperature T of the fuel cell stack 10 , the pressures p in the anode and / or cathode compartment 12 . 13 , the state of charge SOC of the energy storage 44 , the speed n of the traction motor 51 and at least one manipulated variable x.

On the basis of the at least one control value x, for example of the accelerator pedal value determined by means of a pedal value sensor, the control unit determines that of the electrical consumers of the vehicle 200 , in particular the traction motor 51 , requested electrical power P _use . In addition, by setting this power request as a stacked-line power, the controller determines that of the accessories of the fuel cell stack 10 in generating P _net = P _use received electric power P _aux . Finally, the control unit provides to the fuel cell stack a power demand of height P _stack = P _net + P _aux to power the at least one consumer in accordance with the power request.

Depending on the fuel cell stack 10 to be asked power demand P _stack determines the control unit 60 From calculations or correspondingly stored maps, the required mass flows or operating pressures of the anode and cathode operating medium and controls the supply of the fuel cell system, for example via the electric motor 34 of the compressor 33 as well as the adjusting means 24 . 38 etc. of the fuel cell system 100 , In addition, the control unit controls 60 the inverter 43 to the traction engine 51 to provide energy, as well as the converter 45 and optionally other transducers to the energy storage 44 to charge or discharge and to supply the consumers connected to the mains with energy.

According to the prior art, between a first time elapsed t _v1 and the current time from a pedal value as a control value by means of a transfer function stored in the control unit, a power request P _{use of} the electric motor 51 determined. With a stacking _net power P _net corresponding to the power request P _use , the control unit has one from the fuel cell stack 10 to be requested stacking power P _stack determined. Due to the inertia of the resource streams, the increase in _stack power P _{stack is} delayed in time from the power request P _use . In addition, due to a local maximum P _aux ^{max of} a power consumption P _{aux of} the ancillary components of the fuel cell system, a local minimum of the available stacking _net power P _{net occurs} , whereby the net power increase P _{net is} additionally delayed from the increase in _stack power P _stack .

3 shows a curve of a pedal value x _j (t) as the manipulated variable curve of a pedal encoder (PWG), the course of a corresponding power request and a predicted control value. At the time t, the control unit determines a gradient of the control value x on the basis of a current control value x (t) and of a last control value x _j (t _v1 ). _j (t, t _v1 ) = (x _j (t) -x _j (t _v1 )) / (t-t _v1 )) and from this for a first future time t _z1 a predicted manipulated value x _j * (t _z1 ) = x _j (t) + x. _j (t, t _v1 ) * (t _z1 -t). Based on the predicted control value, the control unit determines a predicted power request P * _use (not shown) at time t.

4 shows the course of a pedal value as a control value variation of a pedal encoder (PWG) and the course of an alternatively predicted control value. The control unit determines the gradients x on the basis of two past values of the manipulated variable x _j (t _vi ). _j1 , x. _j2 and x. _{j3 of} the control _value and the gradient x based on a current control _value x _j (t) and a last control _value x _j (t). _{j4 of} the control _value . The control unit determines a predicted manipulated variable for a future time t _z1 according to x _j * (t _z1 ) = x _j (t) + (F1 * _x.j1 + F2 * _x.j2 + F3 * _x.j3 + F4) at time t * x. _j4 ) * (t _z1 - t). The course of the control value during several past Time segments are considered and, according to the weighting factors F1, F2, F3 and F4, different degrees of influence in the prediction of the manipulated variable x _j * (t _z1 ).

5 shows the course of a pedal value as Stellwertverlauf of a pedal encoder (PWG), the course of a corresponding power request and the course of a predicted power request according to an embodiment of the invention. On the basis of how in the 3 or the 4 shown calculated predicted control value x _j * (t _z1) the control unit determines the time t a predicted power demand P * _use for time t _z1. Thus, the power _request P * _use expected for the time t _z1 is already present at the time t and the control unit determines a predicted stack power in a next step (not shown) with P * _use = P * _net and P * _aux = f (P * _net ) P * _stack . If the control unit requests the predicted stack power P * _stack from the fuel cell _stack already at time t, a predicted _stack power P * _net at the time t _z1 corresponding to the predicted power request P * _use is available _without delay. If less power than P * _{use is} requested at time t _zl in contrast to the prognosis, the excess power provided on the basis of the predicted power _request becomes energy storage 44 fed or other consumers, such as a climate system supplied.

6 12 shows the course of a power request, a predicted power request according to an embodiment of the invention, a stack performance, a stacking net power, and a power consumption of at least one accessory when the power request is increased. At time t, a power request P _use increases significantly under control of the control unit. It is irrelevant for the embodiment explained here whether the power request P _use fl ows up due to a current control value change, for example a change in the pedal value, or whether the power request is applied to one as described with reference to FIG 5 explains that at the time t, certain predicted power request P * _{use is} boosted for a future time t _z1, not shown. In both cases, the control unit determines from a map a predicted maximum power consumption P _aux ^max , which is to be expected when the current _stack power P _stack (t) is raised or increased or predicted _stack power P _stack with the highly contended power request P _use = P _net . From this, the control unit determines a transition _{stack power} P * _use (t _im3 ) _such that P * _net (t _im3 ) = P _aux ^max. This step repeats the control unit recursively and thus determines in a next step a second predicted maximum power consumption P _aux ^{max 2} which at increasing the current _stack power P _stack (t) to an increased _stack power P _stack with the transition _{stack power} P * _use (t _im3 ) = P _net . From this, the _controller determines a second transition _{stack power} P * _use (t _im2 ) _such that P * _net (t _im2 ) = P _aux ^{max 2} . Finally, the control unit determines a third predicted maximum power consumption P _aux ^max3 which is to be expected on increasing the current _stack power P _stack (t) to an increased _stack power P _stack with the transition _{stack power} P * _use (t _im2 ) = P _net . From this, the _controller determines a second transition _{stack power} P * _use (t _im1 ) _such that P * _net (t _im1 ) = P _aux ^max3 . Finally, in a first step and at a first time t _im1 , the control unit increases the current stacking power P _stack (t) to P * _use (t _im1 ), increases the stacking power further to P * in a second step and at a second time t _{im2 use} (t _im2 ) and finally increases the stacking power to P * _use (t _im3 ) in a third step and at a third time t _im3 .

LIST OF REFERENCE NUMBERS

100

The fuel cell system

200

vehicle

10

fuel cell stack

11

single cell

12

anode chamber

13

cathode space

14

Polymer electrolyte membrane

15

bipolar

20

anode supply

21

Anode supply path

22

Anode exhaust gas path

23

fuel tank

24

actuating means

25

Brennstoffrezirkulationsleitung

26

recirculation conveyor

30

cathode supply

31

Cathode supply path

32

Cathode exhaust path

33

compressor

34

electric motor

35

power electronics

36

turbine

37

Waste gate line

38

actuating means

39

humidifier

40

electrical power system

41

voltage sensor

42

current sensor

43

inverter

44

energy storage

45

DC converter

50

drive system

51

traction engine

52

drive axle

53

drive wheels

60

control unit

Claims (10)

Vehicle ( 200 ), comprising a fuel cell system ( 100 ) with a fuel cell stack ( 10 ) for providing an electric power P _stack , at least one ancillary unit ( 24 . 26 . 33 . 34 . 38 ) with electrical power consumption P _aux for operating the fuel cell stack ( 10 ) and a net power P _net = P _stack - P _aux ; at least one consumer ( 44 . 51 ) with an electric power request P _use dependent on a manipulated variable x (t); and a control unit ( 60 ); characterized in that the control unit ( 60 ) is set up to ascertain a progression of the detected manipulated _variable and thus a predicted manipulated _variable x * (t _zl ) from values of the _{manipulated variable} x (t) acquired over a specific period of time; from the predicted control _value x * (t _zl ) a predicted power _request P * / use and from this a predicted stack performance P * / stack to investigate; and a current stack performance P _stack (t) by ΔP _stack = P * / stack - P _stack (t) to increase. Vehicle ( 200 ), comprising a fuel cell system ( 100 ) with a fuel cell stack ( 10 ) for providing an electric power P _stack , at least one ancillary unit ( 24 . 26 . 33 . 34 . 38 ) with electrical power consumption P _aux for operating the fuel cell stack ( 10 ) and a net power P _net = P _stack - P _aux ; at least one consumer ( 44 . 51 ) with an electric power request P _use dependent on a manipulated variable x (t); and a control unit ( 60 ); characterized in that the control unit ( 60 ) is adapted to increase a predicted maximum power consumption by ΔP _stack before increasing a current _stack power P _stack (t) P max / aux of the at least one ancillary unit ( 24 . 26 . 33 . 34 . 38 ) when increasing the stacking power by ΔP _stack and, with P max / aux + P _net (t) = P in / net and P in / aux = f (P in / net) , a transitional stack performance P in / stack = P in / net + P in / aux to investigate; and the current _stack power P _stack (t) first P in / stack and then around ΔP in / stack = ΔP _stack + P _stack (t) - P in / stack to increase. Vehicle ( 200 ) according to claim 1 or 2, characterized in that the manipulated variable x (t) is a driver-specific variable, in particular a pedal angle or a pedal angle gradient. Vehicle ( 200 ) according to one of the preceding claims, characterized in that the control unit ( 60 ) is further adapted to the forecasted performance request P * / use as a function of the predicted manipulated _variable x * (t _zl ) and as a function of at least one current vehicle _parameter and / or at least one predicted vehicle _parameter . Vehicle ( 200 ) according to claim 4, characterized in that the at least one consumer an electric motor ( 51 ) and the current vehicle parameter has a rotational speed of the electric motor ( 51 ) and / or the current vehicle parameter is a wheel speed, a vehicle speed and / or a vehicle acceleration. Vehicle ( 200 ) according to claim 4 or 5, characterized in that the at least one predicted vehicle parameter as a function of a gradient of detected values of the vehicle parameter and / or based on current and / or predicted environmental data, in particular a current or predicted road gradient and / or a current or predicted curve radius , is determined. Vehicle ( 200 ) according to one of the preceding claims, characterized in that the control unit ( 60 ) is further adapted to detect a plurality of time-dependent manipulated variable x _j (t) over a certain period Δt = {t _vi , t _{vi + 1} , ..., t _{v (i + k)} , ..., t} and at least one gradient

or

the at least one detected control _value x (t) and for at least one future time t _zl at least one predicted control _value x _j * (t, t _zl ) = x (t) + x. _j (t, t _vi ) * (t _zl - t) or x _j * (t, t _zl ) = x _j (t) + x. _j (t _{v (i + k)} , t _vi ) * (t _zl - t). Vehicle ( 200 ) according to claim 7, characterized in that the control unit ( 60 ) is further arranged for the at least one past time t _vi and / or t _{v (i + k)} and the at least one future time t _zl depending on an operating mode of the vehicle ( 200 ) and / or an operating time of the fuel cell stack ( 100 ) and / or the vehicle ( 200 ) to choose. Vehicle ( 200 ) according to claim 7 or 8, characterized in that the control unit ( 60 ) is further arranged for, depending on the at least one predicted manipulated _variable x _j * (t, t _zl ) at least one predicted power _request P * / use (x _j * (t, t _zl )) and with P * / use (xj * (t, t _zl ) = P * / net (i, j, l) and P * / aux (i, j, l) = f (P * / net (i, j, l)) , at least one forecast stack performance P * / stack (i, j, l) = P * / net (i, j, l) + P * / aux (i, j, l) to investigate; and a current stacking power P _stack (t)

to increase. Method for operating a vehicle ( 200 ), comprising a fuel cell system ( 100 ) with a fuel cell stack ( 10 ) for providing an electric power P _stack at least one accessory ( 24 . 26 . 33 . 34 . 38 ) with electrical power consumption P _aux for operating the fuel cell stack ( 10 ) and a net power P _net = P _stack P _aux ; at least one consumer ( 44 . 51 ) with an electric power request P _use dependent on a manipulated variable x (t); and a control unit ( 60 ); the method comprising the following method steps: detecting the manipulated value x (t _vi ) over a certain period Δt = {t _vi , t _{vi + 1} , ..., t _{v (i + k)} , ..., t}, Determining a course of the detected control _value and thus determining a predicted control _value x * (t _zl ); Determining a predicted power _request P _use as a function of the at least one predicted _{power value} x * (t _zl ); Determine a forecast stack performance P * / stack depending on the predicted service request P * / use; and increasing a current stack performance P _stack (t) by ΔP _stack = P * / stack - P _stack (t).

Images