EP3244046A1 - Control method for a combustion engine, control device and combustion engine - Google Patents

Abstract

The present invention relates to a control method for an internal combustion engine (2) in a vehicle, comprising: determining a command variable (x (t)) taking into consideration operating state information (FW, SB) and a difference (') between an emission upper limit (EM G) and a cumulative actual emission quantity (EM K), influencing an operating state of the internal combustion engine (2) by means of the reference variable (x (t)), so that at least two actual emission quantities are set so that the corresponding cumulated actual emission quantities in an operating period a set of any randomly set different operating conditions of the internal combustion engine (2) shall not exceed emission upper limits (EM G) for that period of operation, minimizing an objective function by using the reference variable (x (t)) ) certain indifference curve (I) from Pareto-optimal age is selected and for the determination of the difference (') a prediction information (PI) is taken into account.

Description

The present invention relates to a control method for an internal combustion engine for determining at least one reference variable for an internal combustion engine.

Important engine functions are set using suitable control methods. Control methods complement constructive measures such as the combustion chamber design and influence the mixture formation in injection systems and injection methods. During engine operation, they reduce fuel consumption and related CO ₂ emissions as well as significant exhaust gas components such as carbon monoxide (CO), hydrocarbons (HC), nitrogen oxides (NOx), and soot and particulate matter.

In this case, information about an operating condition of the engine (for example, speed, torque, desired torque, temperature, DPF (Diesel Particulate Filter) load) is evaluated and set variables that affect the consumption and emissions during operation.

To determine these reference variables are often used in a control method exporting control device additionally stored engine maps in which, for example, a target exhaust gas recirculation rate or a target boost pressure are deposited in dependence on the above operating condition.

Suitable reference variables are, for example, the exhaust gas recirculation rate, exhaust gas recirculation distribution, filling, injection time, ignition point. From these reference variables, manipulated variables are then derived (for example throttle position, position of a VTG (variable turbine geometry)).

The term "internal combustion engine" in this context includes the complete combustion engine system with all its units, auxiliary units and actuators.

This strategy can be used to ensure that emission limits are not met in defined speed profiles by optimizing the allocation of certain reference variables be crossed, be exceeded, be passed. An example of such speed profiles are standardized driving cycles, for example the NEDC (New European Driving Cycle), which are used to determine the exhaust gas and / or consumption values. For such cycles, for example, global optimization approaches are known, as indicated in Heiko Sequence: Emission Modeling and Model-Based Optimization of the Engine Control, D17 Darmstadt Dissertations 2012.

In real driving now occur any different speed profiles and operating conditions that are not known before and during the journey. Since the individual operating states also have different emission values regardless of the engine control, the fuel consumption and emission values (l / 100 km or mg / km) can in some cases deviate significantly below or upwards for these arbitrary, different driving profiles. A global optimization of, for example, fuel consumption or CO ₂ emissions when not exceeding emission limits is thus no longer given by the known control methods.

Particularly in the case of competing emission quantities, such as those that occur in soot emissions (particulate matter) and nitrogen oxide emissions in a diesel engine, situations may arise in which, for example, the permissible nitrogen oxide emissions are exceeded in a speed profile and the permissible soot emissions are significantly exceeded.

Control methods are therefore also in real driving the reference variables - for example, exhaust gas recirculation rate (EGR rate), EGR distribution (high pressure / low pressure), filling, rail pressure, etc. - set optimized but also the use of exhaust aftertreatment systems such as diesel particulate filter and SCR (selective catalytic reduction ) with regard to fuel and AdBlue consumption and emissions.

One possible approach would be to determine a command variable (eg, EGR rate, EGR split, fill), taking into account operating state information, emission caps, and cumulative actual emissions magnitude, which is output to the engine.

The operating state information could include, for example, the speed, the current torque, the desired torque, temperatures, the DPF load, and other variables.

The cumulated actual emission quantity comprises the sum of all emissions emitted by the internal combustion engine during a specific period of operation.

At least one operating state of the internal combustion engine could then be adjusted via the reference variable (s) in such a way that a number of actual emission quantities would be influenced in such a way that the cumulated actual emission quantities in a specific operating period with a combination of any, set in random order, different operating states Combustion engine emission limits for this period of operation would not be exceeded (mg / km) and an objective function would be reduced as much as possible.

Here, a variable to be minimized or optimized is referred to as a target function (eg fuel consumption or the CO ₂ emissions dependent thereon, regeneration intervals of various exhaust aftertreatment systems such as soot particle filters, AdBlue consumption, NOx emissions, etc. or a combination of such variables).

The term "arbitrary" operating states should include all technically meaningful operating states that may occur in the proper normal operation of an internal combustion engine.

Such a control concept would have the advantage that, for example, an uncritical actual emission quantity would be increased by a change in the reference variable so that a critical actual emission quantity would be reduced so as to ensure that the emission limit level (emission limit value) of an emission variable for the critical Emission level in a given operating period would not be exceeded.

In this case, one or more reference variable (s) could be selected by an indifference curve from Pareto-optimal alternatives - from, for example, injection quantity, actual emissions and / or AdBlue dosage. This is done according to a heuristic that takes into account the distances of the cumulated actual emissions to their limit level. The reference variable is therefore determined and adapted dynamically and situation-dependent in this method.

However, such an approach would be reduced to the consideration of past operating conditions.

From hybrid vehicle operation approaches are known in which the torque or power distribution between the engine and the electric motor is optimized taking into account expected driving conditions. Lin et al. A stochastic control strategy for hybrid electric vehicles "For example, sets a stochastic dynamic programming method wherein the optimized driving management strategy for a group of random cycles and is implemented in real time. For example, a dynamic programming is global optimal, but requires a very high computational effort under circumstances so that such control methods in typical vehicle control devices with limited computing capacity may be limited.

It is therefore the object to provide a control method for an internal combustion engine in a vehicle, in which a current command variable is determined in a simple and efficient manner in which expected future driving conditions can be taken into account.

This object is achieved by the inventive control method according to claim 1, a control device according to claim 12 and an internal combustion engine according to claim 13 and a vehicle according to claim 14.

Further advantageous embodiments of the invention will become apparent from the subclaims and the following description of preferred embodiments of the present invention.

The invention is characterized in that a target function is minimized by taking into account a difference between an emission upper limit and a cumulated actual emission quantity when determining the reference variable. In this case, an objective function (eg an emission variable such as the CO ₂ emission, the NO x emission and / or the soot or particle emission) is minimized by selecting the reference variable by means of an indifference curve determined from the difference from pareto-optimal alternatives. To determine this difference, a prediction information is additionally taken into account.

Thus, the method determines the desired reference variables such as exhaust gas recirculation rate (EGR rate), boost pressure / filling, AdBlue dosage or the Distribution of torque or power in hybrid vehicles between electric motor and combustion engine as a function of previous emissions (past analysis) and predicted emissions (prediction information), which are included in the difference analysis. The method is thus based on a computationally easy-to-handle difference analysis, in which both cumulative actual emission quantities are considered as well as prediction information.

Predicted emissions for expected operating conditions are determined, estimated or derived from stored data. For this purpose, for example, route preview information can serve for a planned driving route, which serve for example information about the altitude profile, speed limits, traffic and traffic light information, as well as information on ambient temperatures or atmospheric pressure conditions.

There are implementations in which the prediction information comprises information about a route (for example, the length of the route), which is then multiplied by a route-related emission limit.

There are also embodiments in which the operating state prediction information comprises at least one information from the following group: route quality, route length and environmental conditions. The route quality (eg, gradients), the route length, and environmental conditions (e.g., sea level elevation) can be used to determine essential operating state information (torque, speed) and power requirements of an internal combustion engine.

There is an embodiment in which the prediction information further comprises alternatively or additionally an emission prediction variable. This makes it possible to consider in the difference analysis both the permissible emission limit value - including a predicted component - as well as the actual and expected emission. This makes it possible to analyze the critical difference between this limit and the expected emissions past and future-oriented and to use it to determine the indifference curve.

There are embodiments in which the operating state information comprises at least one rotational speed (n) and a nominal torque (M).

In one embodiment, the actual emission quantities (target functions) comprise at least two of the following variables. The variables include NOx emissions, HC emissions, CO emissions, CO ₂ emissions, combined HC and NO x emissions, soot particle numbers, soot particle mass, charge state of a diesel particulate filter and / or a NO x storage catalytic converter.

In another embodiment, the command variable includes at least one of the following variables that affect emissions behavior, namely, EGR rate, EGR split, fill, spark timing. The manipulated variables derived from this include one of the following variables, via which the desired reference variable can be effected in modern engines, namely throttle valve position; Adjustment of the variable turbine geometry, injection timing, camshaft adjustment.

In another embodiment, two actual emission quantities are considered, in particular the nitrogen oxide output and the soot output, which are competitively related in diesel engines.

There are also versions in which the CO ₂ emissions and the NOx emissions are competitively optimized.

There are also versions in which the CO ₂ emissions, the NOx emissions and the soot emissions, ie three actual emission quantities, are competitively optimized.

With the aid of an internal combustion engine with a control device according to the invention, improved consumption values and emission values can be realized. Such an internal combustion engine is particularly suitable for vehicles.

Fig. 1

schematisch ein Motorsystem mit einem erfindungsgemäßen Steuergerät;

Fig. 2

ein schematisches Fahrzeuglayout mit einem erfindungsgemäßen Steuergerät;

Fig. 3

eine schematische Darstellung eines erfindungsgemäßen Steuerverfahrens mit wesentlichen Input- und Output-Größen;

Fig. 4

ein Diagramm, in dem Ruß- und NOx-Emissionen in Abhängigkeit der AGR-Rate dargestellt sind;

Fig. 5

pareto-optimale Arbeitspunkte, für die eine bestimmte Rußemission und eine bestimmte NOx-Emission gilt;

Fig. 6

Auswahl einer Führungsgröße durch eine Indifferenzkurve basierend auf dem Zusammenhang von Rußemissionen und NOx-Emissionen bei einer bestimmten (erhöhten) kumulierten NOx-Emission;

Fig. 7

die in Fig. 6 dargestellte Auswahl für eine niedrigere kumulierte NOx-Emission;

Fig. 8

die in Fig. 6 dargestellte Auswahl für eine überhöhte kumulierte NOx-Emission;

Fig. 9

die in Fig. 6 dargestellte Auswahl basierend auf dem Zusammenhang von CO₂- und NOx-Emissionen;

Fig. 10

die in Fig. 6 dargestellte Auswahl durch eine nichtlineare Indifferenzkurve;

Fig. 11

eine Darstellung verschiedener Emissionsgrößen über einen Verlauf

Fig. 12

den Verlauf einer Kennlinie zur Bestimmung einer Indifferenzkurve

Fig. 12A-C

unterschiedliche Indifferenzkurven, die nach dem erfindungsgemäßen Verfahren bestimmt sind.

Embodiments of the invention will now be described by way of example and with reference to the accompanying drawings. It shows:

Fig. 1

schematically an engine system with a control device according to the invention;

Fig. 2

a schematic vehicle layout with a control device according to the invention;

Fig. 3

a schematic representation of a control method according to the invention with essential input and output variables;

Fig. 4

a graph showing soot and NOx emissions as a function of EGR rate;

Fig. 5

Pareto-optimal operating points for which a certain soot emission and NOx emission are considered;

Fig. 6

Selection of a reference variable by an indifference curve based on the relationship between soot emissions and NOx emissions at a given (increased) accumulated NOx emission;

Fig. 7

in the Fig. 6 illustrated selection for a lower accumulated NOx emission;

Fig. 8

in the Fig. 6 illustrated selection for excessive accumulated NOx emission;

Fig. 9

in the Fig. 6 illustrated selection based on the relationship between CO ₂ and NO x emissions;

Fig. 10

in the Fig. 6 illustrated selection by a non-linear indifference curve;

Fig. 11

a representation of different emission quantities over a course

Fig. 12

the course of a characteristic curve for determining an indifference curve

Figs. 12A-C

different indifference curves, which are determined by the method according to the invention.

In Fig. 1 a motor diagram is shown, which is controlled or controlled by a control device 1 according to the invention. Shown is as a reciprocating engine 2 (diesel or gasoline engine), trained internal combustion engine, which is filled via valves 3 and a charge air line 4 and is emptied via an exhaust line 5. The supply air passes through an air filter 6 and an exhaust gas turbocharger 7 with adjustable turbine geometry through an intercooler 8 via an inlet valve 3 into the cylinder 9, where optionally via an injection system fuel is supplied. After the compression and combustion of the air-fuel mixture, the resulting exhaust gas is discharged through an exhaust valve 3 via the exhaust line.

The compressed exhaust gas passes through the exhaust gas turbocharger 7, drives it and thus compresses the charge air. Subsequently, it passes through a nitrogen storage catalyst 10 and a diesel particulate filter 11 and finally passes through an exhaust flap 12 into the exhaust 13th

The valves 3 are driven by an adjustable camshaft 14. The adjustment takes place via a camshaft adjusting device 15, which can be activated by the control unit 1.

A portion of the exhaust gas can be introduced via a high-pressure exhaust gas recirculation valve 16 into the charge air line 4. An exhaust-treated partial flow can be conducted in the low-pressure region downstream of the exhaust gas turbocharger 7 via a corresponding exhaust gas cooling 17 and an exhaust gas recirculation low-pressure valve 18 into the charge air line 4. The turbine geometry of the exhaust gas turbocharger 7 is adjustable via an adjusting device 19. The charge air supply ("gas") is controlled via the main throttle valve 20.

About the control unit 1 u.a. the exhaust gas recirculation low-pressure valve 18, the actuator 19, the main throttle valve 20, the exhaust gas recirculation high-pressure valve 16, the camshaft adjusting device 15 and the exhaust valve 12 can be driven (solid lines).

Furthermore, the control unit 1 is supplied via sensors and setpoint generator, for example with temperature information (intercooler 8, exhaust gas cooling 17) and with actual emission values (for example from a sensor or physical / empirical model).

For this purpose, further operating state information can come as: accelerator pedal position, throttle position, air mass, battery voltage, engine temperature, crankshaft speed and top dead center, gear stage, vehicle speed.

So there is a complex control system, which set the engine operation in different operating conditions with respect to different targets, regulate and optimize as possible.

FIG. 2 shows a schematically illustrated vehicle 200 in which the reciprocating engine 2 is arranged with the exhaust line 5 and which is connected via a coupling 24 to a drive train 25.

Optionally or alternatively, the vehicle is provided with an electric drive 23, which is coupled via the clutch 24 to the reciprocating engine 2 or the transmission 2a and the drive train 25. The electric drive 23 is, for example, designed as a permanent magnet synchronous machine which is supplied with energy via an electrical energy store 21 (and a converter 22). The control unit 1 is likewise coupled to the electric drive units (21, 22, 23) via corresponding signal lines (not shown).

The following exemplary embodiments relate to the control and regulation of emission values as a function of predetermined emission upper limits and cumulated actual values.

A basic system for carrying out such a method is in Fig. 2 shown. In this case, the control unit 1 determines one or more required and effective for influencing the emissions and effective reference variables x (t).

The driver's desired value FW, which is derived, for example, via the position of an accelerator pedal and / or a brake pedal, and further operating conditions SB of the vehicle 200 and the engine 2, respectively, also take into account the emission limit values EM ^G which did not exceed during operation and finally a prediction information PI serves to consider future operating states. Typical prediction information PI is, for example, an emission prognosis EM ^P or operating state prognosis information, which, for example, in a vehicle includes information about the route length s (t), the route quality and expected environmental conditions during operation.

From this, reference variables x (t) (eg EGR rate, EGR division, filling, ignition time) are derived and manipulated variables determined in internal combustion engine 2 or its components (for example position of main throttle valve 20, camshaft adjustment, adjustment of turbine geometry of exhaust gas turbocharger 7 , Setting the exhaust valve 12, etc.) affect the emissions (for example, NOx, HC, CO, soot) of the internal combustion engine. These are recorded as mass flows (emission rates) EM ^DS (for example mass per time [mg / s]). From these emissions, cumulated actual values EM ^K of emissions are derived (integration of emission rates over time).

From these cumulated actual values EM ^K , in the control unit 1, together with the elapsed operating time t or the traveled distance s, known or predetermined emission upper limits EM ^G and information about the driver's request FW (eg acceleration: a ^setpoint , torque: M ^setpoint ) and other operating conditions SB (eg speed: v; rotational speed: n) of the internal combustion engine 2 determines the reference variable (s) x (t).

Figure 4 shows by way of example the relationship between NOx emissions and soot emissions as a function of the exhaust gas recirculation rate (EGR), which here forms a reference variable x (t). The graph shows that increasing EGR can reduce NOx emissions but increase soot emissions.

Fig. 5 shows a graph of target size combinations of certain soot emissions plotted against certain NOx emissions. If, for example, there is the task of minimizing / reducing soot emissions in an (arbitrary) operating state, but adhering to a (cumulative) NOx limit value, the emission history (accumulated actual values Em ^G ) for past (possibly arbitrary, in random order set, different operating conditions) are taken into account.

Pareto-optimal target size combinations, in which the soot emission can only be further reduced if the NOx emission is increased, are marked by the points x. All pareto-optimal target size combinations form the so-called pareto front, which connects the points x together. In a minimization problem, points to the left below the Pareto front (shaded area) are unrealizable and all right above target size combinations are not Pareto optimal, since there are combinations (points x) for both soot emission and NOx emission cheaper on the Pareto front can be realized.

The selection from Pareto-optimal target size combinations of two target values (NOx emissions and soot emissions) is shown in the figure in Fig. 6 , In the right-hand column, the emission upper limit EM ^{G is} a NOx limit value NOx-G (dashed line), and the column shown below shows in the hatched area the cumulative actual value Em ^K, the previous accumulated NOx emissions NOx-K ₁ . Since the cumulative NOx emissions NOx-K _{1 are} already relatively close to the NOx limit NOx-G, here a relatively high exchange ratio between the target soot emissions and NOx emissions selected (increased soot emissions in favor of low NOx) to the NOx Limit NOx-G not To exceed. This exchange rate desired here is indicated by the indifference curve I, which is represented here as steeply sloping, and is then shifted to the closest target size combination in which a specific soot emission and a specific NOx emission can be realized for this operating point. This target size combination is then using the in the diagram Fig. 4 known information associated with an EGR as a suitable pareto-optimized command variable x (t).

Fig. 7 shows an example in which the cumulative NOx emissions (NOx-K ₂ ) are further below the NOx-limit NOx-G. Here, the exchange ratio of the indifference curve I is smaller (the straight line is flatter). Here, therefore, a higher NOx emission can be accepted without the risk that the NOx limit value NOx-G is exceeded. Thus, the soot emission can be kept lower. The flattening straight line is shifted to the next target size combination, at which a certain NOx emission and a corresponding soot emission with an associated reference variable x (t) (here the corresponding EGR off Figure 3 ) is feasible.

Fig. 8 shows an example in which the cumulative NOx emissions (NOx-K ₃ ) have exceeded the NOx limit NOx-G. Here, the exchange ratio of the straight line I (vertical indifference curve) is virtually infinite. Regardless of the level of soot emissions, the minimum NOx emission command variable x (t) is selected.

Fig. 9 shows analogously to Fig. 5 an example in which CO _{2 is to be} minimized as a function of the cumulative NOx emissions.

Fig. 10 shows analogously to Fig. 5 an example in which the indifference curve is not linear.

The FIGS. 11 to 12C By way of example, the determination of the indifference curves I (based on the combined emission consideration of the CO ₂ emission and of the NO x emission Figs. 12A-C ), which are applied to different Pareto fronts f, to determine optimized operating points in terms of CO ₂ emissions ṁ _CO2 and NOx emissions ṁ _NOx and derive from it in a known manner, the appropriate (s) reference variable (s).

The basis is the in FIG. 12 shown size β, which is determined depending on a difference δ and corresponds to the slope of the Indifferenzkurven I. φ (β) here denotes the angle at which the indifference curve I intersects the ṁ _NOx axis (abscissa).

δ results according to FIG. 11 from the difference between a time-dependent (or distance-dependent) limit value profile EM ^G over time t (dashed function). The actual limit value EM ^G is thus given, for example, in mg / km, that is to say one mass unit per distance and thus increases with increasing time or traveled distance s. In addition, the course of the cumulative emission values EM ^K (eg a NOx amount m _{NOx is} recorded (solid line) and the difference δ is formed from both (dotted line). $$\delta \left(t\right)=\mathrm{<figure-callout\; id="0"\; label="Max"\; filenames="imgf0002.png,imgf0006.png"\; state="\{\{state\}\}">Max</figure-callout>}\left(0.{\mathit{EM}}_{\mathit{NOx}}^{\mathit{G}}\cdot \left(s\left(t\right)+\tilde{s}\left(t\right)\right)-\left(m_{\mathit{NOx}}\left(t\right)+{\tilde{m}}_{\mathit{NOx}}\left(t\right)\right)\right)$$

Accordingly, the time- or distance-dependent δ results from the emission upper limit EM ^G , for NOx which is multiplied by a route value s, which results from a previous, ie already traveled route s (t) and a predicted route s (t). In the course of the emission cap, therefore, both an information based on actual sizes (previous route) and a route based on predicted information flows in.

The actual, cumulative emission EM ^K (here m _NOx (t)) and a future, predicted emission EM ^P (here m _NOx (t)) are then subtracted from this limit or time-dependent limit value. The prediction is possible up to a prediction horizon t = T.

At a time t = t ₁ , the curves in a retrospective view (arrow V to the left) show the previous curves of the components EM ^G s (t) and m _{NO x} (t), with a forward-looking consideration (arrow Z to the right) additionally the Prediction components EM ^G s (t) and m considers _NOx (t).

mit der abgefahrenen Strecke: $$s\left(t_1\right)={\int }_0^{t_1}v\left(\tau \right)\mathit{d\tau }$$

der prognostizierten Strecke: $$\tilde{s}\left(t_1\right)={\int }_{t_1}^T\tilde{v}\left(\tau \right)\mathit{d\tau }$$

dem kumulierten (bisher realisierten) Emissionswert EM^K, hier: $$m_{\mathit{NOx}}\left(t_1\right)={\int }_0^{t_1}{\dot{m}}_{\mathit{NOx}}\left(\tau \right)\mathit{d\tau }$$

und dem prognostizierten Emissionswert EM^P, hier: $${\tilde{m}}_{\mathit{NOx}}\left(t_1\right)={\int }_{t_1}^T{\tilde{\dot{m}}}_{\mathit{NOx}}\left(\tau \right)\mathit{d\tau }$$

wobei ν̃ die Prädiktion eines zukünftigen Geschwindigkeitsverlaufs ist und ${\tilde{\dot{m}}}_{\mathit{NOx}}$

der damit korrespondierende prognostizierte Verlauf eines NOx-Massenstroms.At the time t = t ₁ then: $$\delta \left({\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="imgf0002.png,imgf0007.png"\; state="\{\{state\}\}">t</figure-callout>}}_1\right)=\mathrm{<figure-callout\; id="0"\; label="Max"\; filenames="imgf0002.png,imgf0006.png"\; state="\{\{state\}\}">Max</figure-callout>}\left(0.{\mathit{EM}}_{\mathit{NOx}}^{\mathit{G}}\cdot \left(s\left({\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="imgf0002.png,imgf0007.png"\; state="\{\{state\}\}">t</figure-callout>}}_1\right)+\tilde{s}\left(t_1\right)\right)-\left({\mathrm{<figure-callout\; id="1"\; label="m\; NOx\; t"\; filenames="imgf0002.png,imgf0007.png"\; state="\{\{state\}\}">m\; </figure-callout>}}_{\mathit{NOx}}\left(t_{}\right)\left({}_1\right)+{\tilde{m}}_{\mathit{<figure-callout\; id="1"\; label="NOx\; t"\; filenames="imgf0002.png,imgf0007.png"\; state="\{\{state\}\}">NOx\; </figure-callout>}}\left(t_{}\right)\left({}_1\right)\right)\right)$$

with the crazy route: $$s\left({\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="imgf0002.png,imgf0007.png"\; state="\{\{state\}\}">t</figure-callout>}}_1\right)={\int }_0^{t_1}v\left(\tau \right)\mathit{d\tau }$$

the forecasted route: $$\tilde{s}\left({\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="imgf0002.png,imgf0007.png"\; state="\{\{state\}\}">t</figure-callout>}}_1\right)={\int }_{t_1}^T\tilde{v}\left(\tau \right)\mathit{d\tau }$$

the cumulative (previously realized) emission value EM ^K , here: $$m_{\mathit{<figure-callout\; id="1"\; label="NOx\; t"\; filenames="imgf0002.png,imgf0007.png"\; state="\{\{state\}\}">NOx\; </figure-callout>}}\left(t_{}\right)\left({}_1\right)={\int }_0^{t_1}{\dot{m}}_{\mathit{NOx}}\left(\tau \right)\mathit{d\tau }$$

and the predicted emission value EM ^P , here: $${\tilde{m}}_{\mathit{<figure-callout\; id="1"\; label="NOx\; t"\; filenames="imgf0002.png,imgf0007.png"\; state="\{\{state\}\}">NOx\; </figure-callout>}}\left(t_{}\right)\left({}_1\right)={\int }_{t_1}^T{\tilde{\dot{m}}}_{\mathit{NOx}}\left(\tau \right)\mathit{d\tau }$$

where ν is the prediction of a future velocity profile and ${\tilde{\dot{m}}}_{\mathit{NOx}}$

the corresponding predicted course of a NOx mass flow.

From a resulting δ value (eg δ1, δ2 or δ3) is then determined by means of a characteristic curve in FIG. 12 is shown, a β value derived, which corresponds to the slope of an indifference curve I, which leads to the determination of a pareto-optimized operating point and thus to the desired reference variable. Pareto fronts f ₁ , f ₂ for different operating states (u _f1 and u _f2 ) are in the Figs. 12A to 12C shown. For the characteristic which leads to the determination of the β-value, it should apply that $${\lim }_{\delta \to 0+}\beta \left(\delta \right)=\infty$$

At the same time, it should also be the case here that the function is strictly monotonically decreasing, so that with increasing δ, β decreases steadily, as is illustrated by way of example in the function in FIG. 13.

In the Figures 12A, 12B and 12C In each case, two curves (the pareto fronts f ₁ and f ₂ ) are drawn, each showing pareto-optimized emission combinations for CO ₂ and NO x in different operating states. Here, f ₁ identifies the emission combinations of an operating state with lower power and overall lower emission values (ṁ _CO2 and ṁ _NOx ) and f ₂ identifies the emission combinations of an operating state with higher power and thus also with rather higher emission values (ṁ _CO2 and ṁ _NOx ). f ₁ and f ₂ form emission parfofronts for different operating states.

The desired reference variable (x (t) for a specific emission _combination u _f1 or u _f2 is determined by applying the indifference curve I whose slope corresponds to the β value resulting from the characteristic curve in FIG. 12 results. This so determined target size combination u _f1 or u _f2 (emission combination) is then assigned, for example by means of a known information AGR as a suitable pareto-optimized command variable x (t)

(analogous to the diagram Fig. 4 ). In FIG. 4 the relationship between NOx and soot emissions is given in relation to the EGR rate. From other diagrams or also from multi-dimensional maps (with pareto surfaces), the relationships between reference variable x (t) and emission combinations from two or more emission quantities can be deduced.

FIG. 12A shows a β ₁ , FIG. 12B a β ₂ and FIG. 12C a β ₃ . The different β-values (β ₁ , β ₂ and β ₃ ) result from the corresponding δ-values with the help of the characteristic curve in the FIG. 12 ,

If the δ between the emission limit value and the cumulative emission decreases from δ1 to δ2, the gradient of the indifference curve must increase (the indifference curve I becomes steeper), since operating points are to be preferred where the NOx emission limit is shorter such operating points are preferred in which the NOx emission is reduced. Accordingly, CO ₂ emissions are increased at these operating points ( FIG. 12B ).

Conversely, with an increasing δ, the β and thus also the slope of the indifference curve, whose course becomes flatter, decrease, and in the desired manner operating points are preferred in which higher NOx values can be tolerated and, on the other hand, the CO ₂ - Emission is reduced accordingly ( FIG. 12C ).

With the presented approach, the emission values (target functions) can be improved during operation and depending on changing boundary conditions. In addition to the problems presented here, in which emission quantities were considered in pairs, the method can also be extended to multi-dimensional problems. It is thus possible, for example, to determine Pareto-optimized reference variables x (t) for multiple combinations (eg for CO ₂ emission, soot emission and NOx emission). In addition to the reference variable AGR, other reference variables x (t) can also be determined pareto-optimized for regulation (eg EGR division, filling, ignition point or rail pressure).

LIST OF REFERENCE NUMBERS

  prognostizierter Verlauf eines NOx-Massenstroms
u_f  Betriebszustand
V  Vergangenheitsbetrachtung
Z  Zukunftsbetrachtung1 control unit
2 reciprocating engine
2a gearbox
3 valves
4 charge air line
5 exhaust system
6 air filters
7 turbocharger
8 intercoolers
9 cylinders
10 NOx storage catalyst
11 diesel particulate filter
12 exhaust flap
13 exhaust
14 camshaft
15 camshaft adjusting device
16 EGR high pressure valve
17 exhaust gas cooling
18 EGR low pressure valve
19 adjusting device
20 main throttle
21 el. Energy storage
22 inverters
23 el. Drive
24 clutch
25 powertrain
200 vehicle
x (t) reference variable
NOx-G limit
NOx-K ₁ accumulated actual value
FW driver's request
SB Other operating conditions
EM ^G emission limit
EM ^C accumulated emission values
EM ^DS emission rates
I indifference curve
PI prediction information
EM ^P emission forecast
δ (t) difference function
β slope
s (t) operating state prediction information
s covered distance
t operating time
f Pareto front
φ (β) angle
s (t) predicted route
ν prediction of a future velocity course
${\tilde{\dot{m}}}_{\mathit{NOx}}$

predicted course of a NOx mass flow
u _f operating state
V Past Consideration
Z Future analysis

Claims (15)

A control method for an internal combustion engine (2) in a vehicle, comprising: Determining a reference variable (x (t)) under consideration an operating state information (FW, SB) and a difference (δ) between an emission upper limit (EM ^G ) and a cumulated actual emission quantity (EM ^K ), Influencing an operating state of the internal combustion engine (2) by means of the reference variable (x (t)), so that at least two actual emission quantities are set so that the corresponding cumulated actual emission quantities in an operating period with a combination of any, set in random order, different operating states of the internal combustion engine (2) emissions upper limits (EM ^G ) for this period of operation, wherein a target function is minimized by the reference variable (x (t)) by means of an indeterminate curve (I) determined by the difference (δ) of pareto-optimal Alternatives is selected and to determine the difference (δ) a prediction information (PI) is taken into account. 2. A control method according to claim 1, wherein the prediction information (PI) comprises operation state prediction information (s (t)). 3. The control method according to claim 1, wherein the operating condition prediction information includes at least one information from the following group: route quality, route length, environmental conditions. 3. A control method according to claim 1 or 2, wherein the prediction information (PI) comprises an emission prediction quantity (EM ^P ). 4. Control method according to claim 1, 2 or 3, wherein the Indifferenzkurve (I) is a straight line whose slope (β (δ)) by means of a difference function (δ (t)) can be determined. 5. Control method according to one of the preceding claims, wherein the target function comprises an actual emission quantity (Em ^DS ), a fuel consumption and / or a CO ₂ emission. A control method according to any one of the preceding claims, wherein the operating state information (SB, FW) comprises a rotational speed (n (t)) and a target torque (Moll (t)). 7. Control method according to one of the preceding claims, wherein the operating period and the different operating states of a ride are known. 8. The control method according to claim 1, wherein the actual emission quantities (Em ^DS ) comprise at least two of the following variables: NOx emission, HC emission, CO emission, CO ₂ emission, combined HC and NOx emission, Soot particle number, soot particle mass, AdBlue consumption. 9. The control method according to claim 1, wherein the reference variable (x (t)) comprises at least one of the following variables: EGR rate, EGR distribution, charge, boost pressure, injection timing, ignition timing, rail pressure. 10. Control method according to one of the preceding claims, wherein at least two actual emission quantities (Em ^DS ), in particular CO ₂ emissions and NOx emissions and / or NOx emissions and Rußausstoß, are considered. A control method according to any one of the preceding claims, wherein at least three actual emission quantities (Em ^DS ) are considered from the following group: CO ₂ emissions, NOx emissions and soot emissions. 12. Control device (1) for carrying out the method according to one of claims 1 to 11. 13. internal combustion engine (2) with a control device (1) according to claim 12. 14. Vehicle with an internal combustion engine (2) according to claim 13.

Images