DE3505485A1 - Method and device for open-loop and/or closed-loop control of the metering of fuel into an internal combustion engine - Google Patents

Abstract

A method and a device for open-loop and/or closed-loop control of the metering of fuel into an internal combustion engine is proposed, in which device an optimum time cycle of the injection is realised on the basis of the overall control concept and in which, by virtue of special correction methods, all the dynamic processes during the injection are also taken into account. Block circuit diagrams of the overall control concept are described and timing diagrams of the timing control are explained. Likewise, a plurality of possible correction methods are illustrated in the form of further timing diagrams. Finally, further possible means are specified which serve to improve the accuracy and the reliability of the described fuel metering according to the invention. <IMAGE>

Description

13.11.198U Sr / Hm '* ""

ROBERT BOSCH GMBH, TOOO STUTTGART 1

Method and device for controlling and / or regulating the fuel metering in an internal combustion engine

ψ- State of the art |

The invention relates to a method and a device for controlling and / or regulating the fuel metering in an internal combustion engine according to the preamble of the main claims.

From the German patent application P 33 OT 826.2 is, in particular from Figure 1, a fuel injection pump known that with the help of an electrically controllable metering valve and an electrically adjustable pump piston is controllable. The electrically controllable metering valve as well as the electrically adjustable pump piston are influenced by an electronic control unit, this control unit of signals with respect to the load of the internal combustion engine, the speed of the internal combustion engine, etc. is applied. All in all

, λ ~ · 35H5485 -X- ; ■ -'--

Thus, in the known device, the fuel metering becomes in an internal combustion engine as a function of the operating parameters characterizing the operating state of the internal combustion engine the internal combustion engine influenced by means of an electronic control unit, and by means of the Fuel injection pump carried out.

It has now been found that the control and / or Regulatory procedure in connection with the described Establishment do not lead to satisfactory results. First of all, it is related to the one described Establishment necessary to improve the previously known overall control concepts, and in doing so before all aspects of the timing and dynamic processes in the known facility described to be considered more precisely.

Advantages of the invention

The method according to the invention and the method according to the invention Device for controlling and / or regulating the fuel metering in an internal combustion engine with the features The main claims have the advantage over the prior art described that with their help in Exact metering of fuel into the internal combustion engine is possible in every operating state of the internal combustion engine is.

This is achieved in that the signals for controlling the metering valve and the pump piston are dependent on the operating parameters of the internal combustion engine are corrected.

■ .jf

- -3Γ-

It is particularly advantageous here to correct the actual or desired operating state of the internal combustion engine to make dependent. Another particularly advantageous development of the invention consists in to simulate the actual operating state of the pump piston, e.g. with the help of a time delay the signal relating to the desired operating state of the pump piston.

Further advantageous developments and improvements of the method specified in the main claims or the specified device are possible through the measures listed in the subclaims.

drawing

Embodiments of the invention are shown in the drawing and explained in more detail in the description below. 1 shows a piston stroke function diagram, FIG. 2 shows a block diagram of the control, FIG. 3 realizations of position control and / or regulation, FIG. H time diagrams for time control and / or regulation, FIG. 5 details of the time control, FIG. 6 a time diagram an alternative time control and / or regulation, FIG. 7 implementation options for the correction, FIG. 8 a time diagram for the third implementation option for the correction, and FIG. 9 a time diagram for the fourth implementation option for the correction.

Description of the exemplary embodiments

The exemplary embodiments described below concern the control and / or regulation of the Fuel metering in an internal combustion engine.

The exemplary embodiments are explained in connection with the diesel internal combustion engine, but this does not mean that the inventive concept on which the exemplary embodiments are based does not also apply to other types of internal combustion engines would be applicable. The exemplary embodiments described below are also not limited to special circuitry Realizations are limited, but can be realized in any embodiment familiar to the person skilled in the art will. It is therefore possible to use the inventive concept on which the exemplary embodiments are based the form of analog circuits, digital circuits, ^ * "with the help of appropriately programmed computing devices, in the form of combinations of these possibilities mentioned, etc. in a corresponding, according to the invention To implement the subject.

· * ·
Figure 1 shows a piston stroke function diagram. In this

^ü Figure 1, a metering valve 10 is arranged so that there is a

Feed line 11 can open or close. In addition, FIG. 1 shows a pressure valve 12 which, in its open state, releases the connection from the fuel injection pump to the injection nozzle. The reference number 13 in FIG. 1 denotes a return memory which consists of a stamp, a spring, etc. in a known manner. Finally, a pump piston bears the identifier 1U, this pump piston lh comprising various grooves, edges and bores. The subject of Figure 1 belongs to the described prior art. More detailed explanations of the subject matter of FIG. 1 can therefore be found in the German patent application P 33 OT 826.2 cited there.

FIG. 1 shows a total of five chronologically successive states of the represented object. In the Figure 1a is the pump piston in its top dead center OT, Figure 1b shows the pump piston 1U in its Downward movement, in Figure 1c there is the pump

penkolben 1U in its bottom dead center UT, while last in Figure 1d and Figure 1e of the pump piston 1U moved upwards, with FIG. 1d showing the start of delivery FB and FIG. 1e showing the end of delivery FE of fuel to the internal combustion engine shows. In the course of time, after the state shown in FIG. 1e, the Top dead center OT of the pump piston 12+ shown in FIG. 1a follow.

In FIG. 1, the single hatched areas are each the base quantity QS, while the double-hatched areas each indicate the quantity QE represent. QK is the injection quantity in FIG. 1, that is to say that which is actually injected into the internal combustion engine Denotes the amount of fuel. Finally, in the figure influencing the metering valve 10 with the aid of the electrical signal ti, as well as influencing the pump piston 1U indicated by means of the electrical signal ae.

If the pump piston 1U is at its top dead center OT, the return memory 13 contains the entire base quantity QS (FIG. 1a). If the pump piston 1U moves downwards, the base quantity QS is again displaced from the return reservoir 13 (FIG. 1b). During the period in which the pump piston \\ is in its bottom dead center UT, the metering valve 10 opens the supply line 11 for a certain period of time dependent on the signal ti, whereby the metered quantity QE is fed to the pump element (FIG. 1c). During its upward movement, the fuel quantity in the pump element, which consists of the base quantity QS and the metered quantity QE, is compressed so that at a certain point in time, namely the start of delivery FB, which is dependent on QS and QE, the pressure valve 12 opens, and as a result, the injection quantity QK is fed to the combustion chamber of the internal combustion engine via the injection nozzle (FIG. 1d). Depending on the angle

'"-"' - / ■ "- ■ 35Q5485

'■■■>

of the pump piston 1 h , thus depending on the signal ae, the pressure valve 12 closes again, and the end of delivery FE is reached (Figure 1e). The base quantity QS remaining in the pump element is then displaced again into the return reservoir 13 in the course of the remaining upward movement of the pump piston 1h , and the entire cycle starts again from the beginning.

In the stationary case, i.e. with constant signals ti and ae, the base quantity QS does not change, es So the following applies: QS = constant. As a result, the injection quantity QK corresponds to that actually injected into the internal combustion engine Fuel quantity exactly the metered quantity QE, it applies so: QK = QE. Since the metered quantity QE is now also directly dependent on the signal ti, it can in the stationary case the fuel quantity QK actually injected into the internal combustion engine can be specified directly by the signal ti. At the same time, in the stationary case, the end of the conveyor FE is directly dependent on the signal ae, and is therefore also unambiguous Are defined.

It is now assumed that the delivery end FE is to be changed with the aid of the signal ae, that is to say that the angular position of the pump piston 1U is to be influenced in such a way that the delivery end FE occurs later, for example. Such a change in the angular position of the pump piston ^ k results in the first moment that the fuel quantity QK actually injected into the internal combustion engine increases, while the base quantity QS consequently decreases to the same extent. This means that if the metered quantity QE continues to be constant, the metered quantity QE and the injection quantity QK are no longer equal. Since it is now desired, however, that the injection quantity QK can be clearly specified in every operating state of the internal combustion engine, the metered quantity QE must be corrected in the described case of a dynamic change in the delivery end FE. This is achieved by

that the desired amount of fuel to be injected into the internal combustion engine during the operating state of a dynamic Changing the fuel metering with the help of a Correction device influenced, and then the metered quantity QE is determined as a function thereof.

One denotes the desired one to be injected into the internal combustion engine Amount of fuel, in short the desired amount with the indicator QW, as well as an amount of fuel as a result the dynamic correction with the identifier QD apply in the steady-state operating state of the internal combustion engine the following relationships:

QK = QW = QE

QD = 0

QS '= constant.

During the period of the operating state of dynamic changes in the fuel metering in the internal combustion engine however, the following relationships apply:

QK = QW

QE = QW + QD

QD φ 0

QS φ constant.

Figure 2 shows a block diagram of the control. In this Figure 2, the two reference numerals 20 and 21 are two functional transformations that differ from one another. The functional transformation 20 is the main one depends on the speed of the internal combustion engine K and the position FP of the accelerator pedal of the internal combustion engine, while the functional transformation 21 mainly of signals relating to the speed of the internal combustion engine N, the temperature of the internal combustion engine or the temperature of the intake air T, as well as, for example, the pressure in the intake pipe

the internal combustion engine or the atmospheric pressure P is applied. The output signal of the functional conversion 20 represents the desired amount of fuel to be injected into the internal combustion engine, ie the desired amount QW. This output signal QW is fed to the functional conversion 21 on the one hand and to a summation 26 on the other. In contrast, the output signal of the functional conversion 21 is the base quantity QS. Depending on this base quantity QS, as well as other operating parameters characterizing the operating state of the internal combustion engine, a position control and / or regulation 22 forms an output signal which represents the setting angle of the pump piston 1U and which is denoted by ae. Finally, the setting mechanism 23 for the pump piston 1U influences the position of the pump piston Ik, and thus the end of delivery FE, as a function of the signal ae. Depending on the operating state of the internal combustion engine, a correction 27 generates an output signal which corresponds to the dynamic correction quantity QD and which is fed to the summation 26. The last-mentioned summation 26 forms an output signal from its two input signals QW and QD which is denoted by QE and which corresponds to the metered quantity QE. This signal QE is connected to a time control and / or regulation 2k , which is also acted upon by operating parameters characterizing the operating state of the internal combustion engine. Depending on its input signals, this time control and / or regulation forms an output signal which determines the injection duration of the metering valve 10 and which is identified by ti. With the aid of an output stage 25 for the metering valve 10, the metering valve 10 is then influenced as a function of the signal ti acting on the output stage. Finally, an injection nozzle is identified in FIG. 2 with the reference number 28, the supply of fuel via this injection nozzle 28 into a combustion chamber of the internal combustion engine being clearly defined by the actually injected fuel quantity QK and

- / -: ■ '' Λ ■ ß- '

through the delivery end FE of the injection. It should be noted that identical designations of the figure and FIG. 2 also have the same meanings.

As already described in connection with FIG. 1 " the "two signals ae and ti do not change in the steady-state operating state of the internal combustion engine." This means that the injection quantity QK and the end of delivery FE also remain constant. As also in context has already been carried out with FIG. 1, however, when changing the setting angle of the pump piston 1U, that is, a change in the signal ae, the injection duration of the metering valve 10, that is to say the signal ti, is corrected will. This is done in that the signal QD, which is equal to zero in the stationary case, with the Signal QW is linked, and thereby the metered quantity QE is changed.

The following applies in all operating states of the internal combustion engine, that is to say in the stationary as well as in the dynamic case the following relationship:

QK = QW = f (N, FP, ...).

The following relationship also exists in all operating states of the internal combustion engine:

FE = f (QS) = f (QW, ...).

Overall, this has the consequence that, in all operating states of the internal combustion engine, those actually of the internal combustion engine injected fuel quantity QK and the delivery end FE of the injection are at least dependent on the speed of the internal combustion engine N, as well as the position FP of the accelerator pedal of the internal combustion engine. Included is, however, in all operating states at the same time

Λ ~ · - · - ■ ty i •

Internal combustion engine a clear relationship between the injection quantity QK, the end of delivery FE and the speed N, as well as the position of the accelerator pedal FP, this relationship with the aid of the functional transformations 20 and 21 can be set. It is particularly advantageous here to have these functional deformations 20 and 21 to be implemented in the form of at least two-dimensional maps.

The position control and / or regulation 22 of FIG. 2 will now be described in the following text with the aid of FIG. 3, the time control and / or regulation 2U with the aid of FIGS. K, 5 and 6, and the correction 27 of the figure with the aid of FIGS. 7> 8 and 9 · For more precise information about these three components of FIG. 2, the corresponding description of the cited figures should therefore be used. The interlocking 23 for the pump piston 1U and the output stage 25 for the metering valve 10 are devices that belong to the prior art and are therefore familiar to the person skilled in the art. These two last-mentioned elements of FIG. 2 are therefore not to be explained in more detail.

FIG. 3 shows possible ways of realizing the position control and / or regulation 22. In FIG. The input signal of this adjustment 31 forms a signal with respect to the actual operating angle of the pump piston lh, indicated with aeist. A control 32 is controlled by the output signal of the summation 30 and, as a function of this, generates a signal ae.

In contrast, in FIG. 3b there is a summation 33 on the one hand from the signal QS and on the other hand from the output signal an adaptation 3 ^ · driven, with the

last-mentioned adaptation 3 ^ with a signal with respect to the actual start of injection SBist is applied. The output signal of the summation 33 is sent to a control system 35 connected, which forms an output signal as a function of this input signal, which is sent to another Summation 36 is connected. As a further input signal If the summation 36 is the output signal of a further adaptation 37 is supplied. The latter Adaptation 37 is dependent on a signal relating to the actual setting angle of the pump piston 1U, whereby this signal is again marked with aeist. The summation 36 combines its two input signals, and then applies the logic signal obtained therefrom to a further control 38. The output signal the control 38 is then the signal ae.

The two signals aeist and SBist appearing in FIGS. 3a and 3b are actual value signals, that is to say signals which reflect the actual state of certain parameters of the fuel injection pump. As already mentioned, the signal aeist is the actual setting angle of the pump piston 1 h. This signal aeist can be obtained, for example, with the aid of an inductive displacement sensor on the control rod which determines the setting angle of the pump piston 11 *. As has also already been mentioned, the signal SBist is a signal relating to the actual start of injection. In this context, the start of injection is understood to mean the point in time at which the injection nozzle 28 in FIG. 1 opens the fuel supply into the combustion chamber of the internal combustion engine. In terms of time, the start of injection is usually a short time after the start of delivery, since the pressure wave generated by the fuel injection pump requires a finite time to get from the fuel injection pump to the injection nozzle via a high-pressure line. The signal SBist can, for example, with the help of an inductive

Injection start sensor, which measures the movement of the nozzle needle in the injection nozzle, are included.

The subject of FIG. 3 a is a position control circuit known per se, which regulates the actual setting angle aeist of the pump piston 1 k to a setpoint value defined with the aid of the signal QS. In contrast, the subject of FIG. 3b has, in addition to the known position control loop, a further, superimposed control loop. With the help of the superimposed control loop, the setpoint value of the position control loop, that is to say the output signal from element 35 of FIG. In connection with FIG. 3b, the signal QS therefore has the meaning of a desired start of injection.

When the position control and / or regulation 22 of FIG. 2 is implemented in accordance with the block diagram of FIG. 3b it is therefore possible not only to regulate the adjustment angle of the pump piston ^ 1U to its desired value, and so that the actual end of delivery is set to a desired value, but rather based on the superimposed control loop the start of injection can be controlled to the desired, optimal values at the same time. This is particularly due to the fact that the start of injection has a great influence on the course of the combustion is particularly advantageous.

FIG. K shows time diagrams for the time control and / or regulation 2k. A jump in the amount of fuel to be injected to smaller values is shown in FIG. 1, while FIG. B shows a jump in the amount of fuel to be injected towards larger values. In both figures, the time axis is plotted at the top, with the time of the load change on this axis,

that is, the point in time of the jump to "smaller" or larger Fuel quantities, marked with the designation tW. Underneath, the development over time of the cam 1 is shown, • this cam 1 to the cylinder 1 and thus also to the Solenoid valve 1 belongs. The term reference denotes a signal that is generated once per camshaft revolution occurs and can therefore be used to distinguish the individual cylinders from one another. In the present two figures Ua and Ub is the Reference signal clearly between those explained below Segment pulses 0 and 1 approximately in a range at the end of the occurrence of cam 1, but independent of the assignment of the cam end to the segment signal. So it is possible using this special reference signal the cam 1 can be clearly recognized. The signal is used to calculate the speed of the internal combustion engine, which is identified in the figure U with the expression segments. In the two timing diagrams of the Figure U, the segment signal has a total of 12 segments per camshaft revolution, twice the number of cylinders. Since "it is therefore in the case of the one described in the exemplary embodiment Internal combustion engine to a six-cylindrical internal combustion engine acts, this has the consequence that with the help of the segment signal, the individual cylinders of the internal combustion engine can be identified because the distance between the individual cylinders is always exactly two segments amounts to. Next is also in the two figures Ua and Ub another 1 ms cycle applied. This clock serves, as will be shown later, the simplified one Time counting. Overall, it is therefore possible with the help of the reference signal and the segment signal, each individual Identify the cylinder of the internal combustion engine and assign a number to it.

■■ 350 & 485

-A

As a next step there is a signal in the two figures Ua and Ub shown, which is designated by MV1, and with which the solenoid valve 1, that is to say that of the cylinder 1 and the cam 1 belonging solenoid valve is activated. If one considers the development of the cam 1 in the two figures Ua and Ub, the time domain can be derived from this, in which the pump piston 1U lingers in its bottom dead center UT, to then be raised to its top dead center OT due to the rise of the cam and then to fall off again. As has already been explained in connection with FIG. 1, the actual injection takes place of fuel into the internal combustion engine, i.e. the time of the start of delivery FB and the end of delivery FE in the time range of the rise of the cam, i.e. shortly before its top dead center OT. For this reason, the Metering of fuel into the pump element take place before the rise of the cam, i.e. in the time range in which the pump piston is in its bottom dead center BDC. This time range includes in this embodiment an angle of 2U0 NW, with the letters NW to indicate that it is is a camshaft angle. If you take manufacturing tolerances, necessary corrections, etc. into account, it remains a time range remaining as a possible metering range which corresponds to a camshaft angle of 195 NW. This The metering range is indicated in the two figures Ua and Ub in connection with the development of the cam 1. In terms of the time sequence, the individual segment pulses of the Segment signals are numbered, starting with 1 for the segment pulse directly after the reference signal. Any odd number Segment pulse represents a reference mark for a specific cylinder, so segment pulse 1 for cylinder 1, segment pulse 3 for cylinder 2, and so on Occurring odd segment pulse the time range of the 195 NW belonging to the respective cylinder is started. For this purpose, however, the said cam must first

wave angle range can be converted into a time period as a function of the speed of the internal combustion engine. Since the previous processes involved cam 1 and thus cylinder 1, it is now known in which time range the solenoid valve MV1 belonging to cylinder 1 can meter fuel into the associated pump element. As shown in the two figures Ua and k ~ b can be seen, this metering time range is divided into a delay time and a quantity pulse time. The quantity pulse time is always at the end of the metering time range. This has the advantage that changes in the desired fuel quantity to be metered can be reacted to as late as possible. For this purpose the quantity pulse time, which corresponds to the injection duration ti of the metering valve 10 and is therefore known, is subtracted from the also known time duration available for the entire metering in order to obtain the delay time as the result the segment signal started, but, as already mentioned e, for easier counting only with the occurrence of the next 1 ms signal following the segment signal. Furthermore, the delay time is rounded up or down to an integer multiple of 1 ms, so that the end of the delay time also coincides exactly with a 1 ms signal. For the delay time, it is therefore only necessary to count in the 1 ms grid selected in the example. In connection with the quantity pulse time, however, such rounding up or down is not permitted due to the requirements for the accuracy of the fuel metering. For this reason, the quantity pulse time is first roughly counted in a 1 ms grid, in order to then precisely count the remaining amount of the total, exact quantity pulse time, e.g. in a U microsecond fine grid. So overall the

'3: 6.05485

Delay time chosen so that it can be counted by means of a coarse grid, whereas the Quantity pulse time is first counted with the same coarse grid, in order to then get the exact value of the quantity pulse time to be achieved by means of a fine grid. The delay time is chosen so that the sum From the delay time and the quantity pulse time, precisely the speed-dependent duration of the described 195 NW Area results.

The signals MV2 to MV6 for controlling the solenoid valves of cylinders 2 to 6, which are also shown in the two figures Ua and k ~b, correspond in principle to the signal MV1. In contrast to the signal MV1, however, the signals MV2 to MV6 differ in the two figures ka and hb. This is due to the fact that in the figure ka less fuel is to be supplied from the point in time tW, while in the figure k ~ b more fuel is desired. The point in time tW is again indicated by an arrow in connection with the signal MV2.

As can be seen from FIG. Ka , the point in time of the load change tW occurs during the quantity pulse time of the signal MV2 and the signal MV3. Since, in connection with FIG. Ka, after the point in time of the load change, less fuel is desired, the quantity pulse time for the two signals MV2 and MV3 is terminated directly after the point in time tW. The termination takes place in the 1 ms coarse grid, whereby, as will be explained later, only one solenoid valve changes its position at a time. According to FIG. Ka , the quantity pulse time of the signal MV2 is therefore terminated directly after the point in time tW in a 1 ms grid, while the signal MV3 ends a further millisecond later. The point in time falls for the solenoid valves of cylinders k, 5 and control signals MV ^, MV5 and MV6

- ι / - Iv ν;

of the load change tW does not affect its quantity pulse time. It is for this reason that it is the case with the latter signals possible to extend the delay time so that the quantity pulse time corresponding to the new fuel quantity requirement comes back to the end of the 195 NW range. Overall, therefore, due to the load change in Time tW a change in the normal sequence, namely the termination, only for the two signals MV2 and MV3 their respective pulse time. This termination can take place as early as desired during the quantity pulse time. Lies the load change before the quantity pulse time, i.e. in the delay time, is the same as with the signals MVU, etc. shown, the delay time is extended so that the now due to the reduced amount of fuel required shorter volume pulse time reaches the end of the 195 NW range again.

As already mentioned, the signals are in the figure Ub to control the solenoid valves of the internal combustion engine in the event of a jump in the desired amount of fuel to be metered shown for larger values at time tW. As can be seen from the figure Ub, the point in time is located tW after the end of the volume pulse time of signal MV1, but during the volume pulse time of the signals MV2 and MV3, as well as before the quantity pulse time of the signal MVU. This has the consequence that already during the metering of fuel in cylinder 2 of the internal combustion engine takes into account a portion of the desired increase in fuel quantity can be. Since now, for reasons that have already been explained, the actual quantity pulse time is already on the end of the 195 ° UW area has been placed, this can actual quantity pulse time can no longer be arbitrarily extended. As a maximum limit for metering of fuel into the pump element is the end of the range of bottom dead center UT, that is

the 2Uo ° NW area shown in connection with the development of cam 1. Like the signal MV2 of the Figure Ub can be seen, therefore, the volume pulse time of the signal MV2 over the end of the 195 ° NW range extended to the limit set by the 2U0 ° NW area. This limit is from the existing, numbered segment signal can be derived. The same extension of the quantity pulse time up to her maximum limit takes place with signal MV3. With the signal MVU, the time tW occurs during the actual delay time so that with this signal it is possible to abort the delay time within the 1 ms grid and to let the quantity pulse time begin as soon as possible after time tW. Because of this termination the delay time and the brought forward Quantity pulse time, it is no longer necessary due to the desired increase in fuel quantity To extend the volume pulse time up to the maximum limit set by the 2Uo NW range, but rather it is sufficient to let the quantity pulse time last only a short time beyond the 195 ° MW range. in the In connection with the time sequence described, reference is again made to the signal MVU in FIG. Ub. In this figure Ub, the signal MVU breaks off its delay time shortly after the point in time tW indicated by the arrow, with which its quantity pulse time begins. While this volume pulse time is then calculated the length of this volume pulse time to finally before the maximum limit to end the volume pulse time according to the calculated value. The signals MV5 and MV6 correspond in principle again to signal MV1, only now, due to the desired increase in fuel quantity the delay time has been shortened and the quantity pulse time has been extended accordingly.

■ Ιλ-

Finally, about the two figures Ua and Ub it should be noted that these show the worst case scenario of high speed and large amount of fuel desired that it It is therefore possible that several signals for controlling the solenoid valves of the internal combustion engine are simultaneously the have the same state, i.e. that e.g. several solenoid valves are open at the same time and thus fuel is metered into the corresponding pump elements. in the In the event of such overlapping of the solenoid valve control signals, it is then entirely from the point in time of a load change depends on how this change in load changes the individual solenoid valve control signals. the Figures Ua and Ub show examples of two possible load state changes at specific times. All possible other load changes then cause corresponding analog changes in the solenoid valve control signals.

In a particularly advantageous manner, it is possible to use the timing diagrams shown in FIG to realize appropriately programmed electronic computing device. For example, the coarse grid of the delay time can be used and the quantity pulse time realized with the help of a counter implemented by programming while the fine grid is counted with the help of a hard-wired counter, for example. It is possible assign a dedicated, programmed counter to each solenoid valve control signal. The individual functions such as opening and closing the solenoid valve, can then be done, for example, with the help of interrupt programs can be implemented, whereby these interrupt programs can be activated by the individual counters. All in all does not represent the implementation of the timing diagrams shown in FIG. U as fundamental for a person skilled in the art Problem, and should therefore not be explained in more detail.

As already mentioned in connection with the description of FIG. K , it is useful or possibly even necessary to avoid high inductive loads, for example, to select the switch-on and switch-off times of the individual solenoid valves so that only a single solenoid valve is switched. Also, when realizing the timing diagrams in FIG. H, for example with the aid of an electronic computing device with appropriately programmed counters, it must be ensured that the counters are always correctly reset, so that under no circumstances does a counter, for example, cause a solenoid valve to open incorrectly.

Figure 5 shows details of the timing. With the help of these additional measures, it is possible to determine the correct input and Ensure that the solenoid valves are switched off safely.

In FIG. 5, two counters implemented, for example by means of programming, are plotted over the time axis, namely on the one hand the counter Z1 belonging to the solenoid valve 1, and thus the cylinder 1, and on the other hand the counter Z2 belonging to the solenoid valve 2 and thus to the cylinder 2. Next, the two solenoid valves 1 and 2 are shown signals driving MV1 and MV2, the k corresponding to the respective signals of FIG. Finally, a signal E2 and a signal A2 are shown over the time axis in FIG. Overall, the two signals MV1 and MV2 of FIG. 5 each show two metering pulses, or, in accordance with FIG. K, each two amount pulse times. The start of such a metering pulse is defined by the end of the delay time VE, whereas the end of the metering pulse is clearly defined by the end of the quantity pulse IE. Since, as I said, there are two solenoid valves and two metering pulses each,

-JSTf- - ■ '■.

indicates the first number after e.g. the abbreviation VE die Identification of the pulse, whereas the second number identifies the solenoid valve. This is also the reason the first metering pulse of the solenoid valve 1 with 111 and the second metering pulse indicated by 121 in FIG. Analog applies to the solenoid valve 2, which has a first metering pulse 112 and a second metering pulse 122.

The mode of operation of the two counters Z1 and Z2 is shown in FIG. 5 with the help of evenly rising straight lines, whereby the value to be counted by the counter is identified by a horizontal threshold. Achieved If the counter exceeds this threshold, this means that it is either the end of the delay time VE or the end of the quantity pulse IE has reached. When the end of the delay time is reached, an interrupt for opening the associated solenoid valve is triggered and the counter is reset at the same time. Reaching the end of the quantity pulse time on the other hand, only triggers an interrupt to close the solenoid valve, but the counter runs unabated continue until the next start of a delay time. From Figure 5 it can now be seen that, for example, the switch-on lock E2 is always active for the solenoid valve 2 when the delay time expires, whereas the Switch-off lock A2 of the solenoid valve 2 then prevents the solenoid valve from being switched off while the Quantity pulse time expires. This prevents the solenoid valve 2 from erroneously opening at time ta is opened, because without the switch-on lock E2, the solenoid valve 2 would be opened at time ta, since the Solenoid valve 1 is open and the counter Z1 has a value greater than the threshold VE12. Analog would Without the switch-off lock A2, the solenoid valve 2 cannot be switched off at time te, since it is at time tb would have already been "switched off". At this point in time tb, the solenoid valve 2 is closed, but since

At the time tb, the solenoid valve 1 is also closed and the counter Z2 has a value greater than that Threshold IE22 the solenoid valve 2 would already be "switched off" at time tb, even if it was not switched on at all so that the actual switch-off would no longer be possible at that point in time. Analog inputs and Switch-off locks also exist for the remaining solenoid valves, so that in this way the individual solenoid valves the internal combustion engine are completely decoupled from one another.

Overall, the switching on of this solenoid valve is therefore due to the switch-on and switch-off lock of a solenoid valve only possible if firstly the previous solenoid valve was opened last, secondly the end of the delay time of the respective solenoid valve has been reached, and thirdly, the counter after the Has been reset again when the end of the delay time has been reached. The switching off of a solenoid valve is analogous only possible if firstly the previous solenoid valve was switched off last, secondly the solenoid valve in question is currently switched on and, thirdly, the end of the quantity pulse has been reached has been.

It should also be pointed out in this connection that the quantity pulse time ends in FIG is the end of the fine counting of the respective quantity pulse time, but this fine counting is not shown in more detail in FIG.

FIG. 6 shows a time diagram of an alternative time control and / or regulation 2k. In this Figure 6, the signals MV1 to MV6 for controlling the solenoid valves of the individual cylinders 1 to 6 are shown over the time axis.

wear. So-called stroke numbers HN on the one hand and a reference on the other hand are entered as special points in time in FIG. In the latter reference is a reference signal, as occurs h in the figure. The stroke numbers, on the other hand, each identify a stroke of a cam, so that in comparison to Figure k , the area of two segment signals always makes up the distance between two stroke numbers. Since, in connection with FIG. 6, this is again a six-cylinder internal combustion engine, after six strokes, that is to say after the stroke number HN = β, all processes are repeated again from the beginning. Instead of the distance from stroke to stroke, the segments or other possible speed pulses can also be used as coarse grids. Furthermore, the two points in time of the end of the delay time and the end of the quantity pulse time are also identified by two arrows in FIG.

With the aid of the reference, each stroke number HN is again uniquely assigned to a cylinder in a manner analogous to FIG. H. In the exemplary embodiment described in FIG. 6, the stroke number HN = 1 corresponds to cylinder 1, and thus also to signal MV1, etc. For the further explanation of the alternative time control and / or regulation 2U, it is necessary to introduce a stroke number HZ. This number of strokes HZ is precisely the number of strokes that the delay time runs through completely. Since exactly one delay time must begin with each stroke number HN in the alternative 'native time control according to FIG. 6, the stroke number HZ is exactly one when the delay time is longer than the distance between two stroke numbers, but shorter than the distance from. three stroke numbers. The entire delay time or the entire quantity pulse time is now made up of the time interval between two or more stroke numbers, the coarse grid and one

Fine counting up to the value of the desired length of time together. Usually, however, the distance between two is sufficient Stroke numbers, so that at the time each stroke number occurs, both counters, i.e. the counter for the end of the delay time, as well as the counter for the quantity pulse time end for the corresponding solenoid valves together can be started. Through the interrupt signals ultimately triggered by the two counters the corresponding solenoid valves are then opened or closed. The selection of the solenoid valves is done with the help of the following relationships. When an interrupt signal arrives due to the end of the delay time VE, the (HN - HZ) th solenoid valve is switched on, while when an interruption signal arrives, due to of the quantity pulse time IE the (HN + Z / 2) th solenoid valve is switched off. With Z is the number denotes the cylinder of the internal combustion engine.

As an example, the point in time of stroke number HN = 3 should be described again in more detail. The timing of the Stroke number HN = 3 is assigned to cylinder 3, and thus • also to signal MV3 or the segment pulse for the cylinder 3 or the solenoid valve 3. At the time HN = 3, the solenoid valve 1 and the solenoid valve 6 are switched on, while the rest of the solenoid valves are off. A delay time is also started at time HN = 3, namely the rough counting of the delay time for the solenoid valve 3, so the delay time of the Signals MV3. As a result, with the previous stroke number HN = 2, the rough counting of the delay time for the signal MV2 must have begun, this delay time not yet ended, since the solenoid valve

ir

2 is still switched off. Furthermore, at the point in time HN = 3, the fine counting of the delay time of the solenoid valve is performed 2 and the volume pulse time of the solenoid valve 6 started. Since now, as has already been stated, normally

the delay time only completely covers the area between two stroke numbers, with a maximum of the area of four stroke numbers is possible, and there is also the ■ delay time of the MV2 signal was started at the time HN = 2, the following relationship normally applies to the number of strokes HZ: HZ = 1. This means that according to the definition (HN -HZ), the second solenoid valve must be switched on. This The process corresponds to the signal MV2 of FIG. 6. According to the definition (HN + Z / 2) must also be used in the example used six-cylinder internal combustion engine, the sixth solenoid valve can be switched off. This too corresponds to the signal MV6 of FIG. 6. Overall results Thus it can be seen from FIG. 6 that the delay time end VE occurring after the point in time HN = 3 for Solenoid valve 2 is associated, while the quantity pulse time end, which also occurs after the point in time HN = 3 IE belongs to solenoid valve β. In an analogous manner, stroke numbers can also be used in all others defined times the affiliation of the delay times started at these times and quantity pulse times for the individual solenoid valves can be determined. It is only necessary to ensure that the two specified relationships in a corresponding way, e.g. by means of a modulo 6 calculation the corresponding internal combustion engine can be adapted have to.

As already mentioned, the delay time usually always covers the same distance between stroke numbers completely, so that the number of strokes HZ is usually always the same. In this operating state of the internal combustion engine, which essentially coincides with the stationary operation of the internal combustion engine, it is not necessary to the assignment of the delay time to be started and the quantity pulse time to be assigned at the time of each stroke number to recalculate the individual solenoid valves, but in this steady state of the internal combustion engine

it is sufficient to relate the delay time and the quantity pulse time to the next solenoid valve. This is due to the fact that, in the steady-state operating state of the internal combustion engine, the solenoid valves are regularly switched on and off cyclically one after the other. A steady-state operating state of an internal combustion engine is shown in FIG. 6 , and the regular switching on and off of the solenoid valves is indicated by the numbers 1 and 0 at the two times HN = 3 and HF = k , with a 1 indicating an activated solenoid valve and an O represents a deactivated solenoid valve. The transition from the point in time HN = 3 to the point in time HN = k can thus be achieved according to FIG. 6 simply and without any calculation by shifting the vertical numerical pattern of the ones and zeros down one place in the circle.

In the transient operating state of the internal combustion engine, however, it is possible that the desired delay time is less than the simple distance between two stroke numbers, or greater than twice the distance between two stroke numbers. In these cases it is necessary to adjust the number of strokes HZ, and from this, according to the two relationships mentioned, to calculate the solenoid valves that will be operated on Must be switched on and off at the end of the delay time or quantity pulse time to be started. Overall is it is thereby possible in the unsteady operating state of the internal combustion engine that displacements and overlaps the delay and quantity pulse times of the individual signals for controlling the solenoid valves occur.

Sudden metering changes can also be taken into account within a rough grid, if necessary existing period counter for counting the time interval between the individual segment pulses of the segment signal by exceeding a predeterminable metering

- .27 -

-Λ-

change threshold "on the fly" can be queried. In this case, sudden changes in metering can be used with With the help of the current metering time obtained, the said metering change can be taken into account.

The alternative time control according to FIG. 6 is mainly a simplified possibility of counting the delay time and the quantity pulse time. The calculation of these times, as well as the corresponding reactions to load changes according to FIG. H, as well as the decoupling of the time control according to FIG. 5, can be applied in a manner analogous to that described in connection with the cited figures. However, it is possible that in the case of the alternative timing according to FIG. Β certain adaptations of the calculations and reactions have to be carried out. It is also possible that, in connection with the alternative control of FIG. 6, the execution of the remaining elements shown in FIG. 2 must also be adapted to this time control. In particular, a change in the functional transformations, that is to say the map values, may be necessary.

The use of segments as a coarse grid is particularly advantageous in the alternative time control according to FIG. 6. The consequence of this is that, for example, when the time control is implemented with the aid of an electronic computing device the e.g. programmed counters for the rough counting can be saved. Also be this way all interruptions that previously occurred due to the rough counting are avoided. With the alternative Time control according to FIG. 6, it is only necessary to have two counters for the fine counting of the delay time and the quantity pulse time to be provided. Consists

the possibility of reading the count of these counters on the fly, i.e. without stopping the counters, is sufficient even a single counter. Another advantage of the alternative timing in this context is the higher one Accuracy of this time control, in particular the position of the metering pulse, since not only the volume pulse time im Fine grid is counted, but also the delay time. So overall it is with the help of the alternatives Time control possible, which is necessary for the control of the solenoid valves of the internal combustion engine Program runtime of an electronic computing device shorten.

As stated in connection with the description of Figures 1 and 2, it is necessary, e.g. Change the delivery end FE the desired, the internal combustion engine to correct the amount of fuel to be supplied QW with the aid of the correction amount QD. During the change of the delivery end FE, the equality of the desired quantity QW and the metered quantity QE is no longer given. Nevertheless, due to the correction amount QD, the equality of the desired amount QW and the actual amount remains the internal combustion engine supplied fuel quantity QK obtain.

FIG. 7 shows possible ways of realizing the correction 27. In this FIG. 7, the designation 50 is a Input signal conversion marked, their output signal is connected to a correction calculation 51. The input signal of the conversion 50 is 11, the output signal is designated by 12. In contrast, the output signal of the correction computation 51 is a signal relating to the corrective amount QD. Finally, the correction computation 51 is also used in addition to the signal 12 acted upon by further input variables characterizing the operating state of the internal combustion engine.

A first possibility for generating the correction quantity QE can be realized with the aid of the following equation will:

QD = K1. aresin (K2. ä ~ ft).

With this first possibility, the correction amount QD is therefore dependent on the mean control rod speed won. Of course, it is also possible in the aforementioned relationship instead of the signal ae the signal must be selected. In connection with the figure T there corresponds to the signal ae or aeist das Signal 11 and the signal ae or ae is the signal 12. The input signal conversion 50 of FIG. 7 thus generates from the control path a signal relating to the mean control path speed, while the correction calculation 51 then performs the remaining calculation according to the above equation.

As a second possibility, the correction amount QD can be calculated according to the following equation:

QD ~ K1.aresin (K3.ae).

In the present equation, the amount of correction becomes dependent on the change in the control path generated during two predetermined times. Of course, it is also possible here instead of that Signal ae also the signal related to the actual Usually mostly to use. The assignment of the signals mentioned to the signals in FIG. T corresponds to this the assignment as it relates to the first Possibility is explained. Finally, to the first and the second realization possibility the FIG. 27 also mentions that the designations K1, K2 and K3 are constant values which are dependent are based on the geometrical data of the fuel injection pump .

As a third possibility of generating the correction amount QD, it is possible to do something similar to the second Possibility already explained to generate the correction quantity QD in that now instead of the signal ae the signal QS is used. This means that the signal QS corresponds to the signal 11 of FIG. 7, that from this signal the conversion 50 generates the signal QS as an output signal, and that the correction calculation 51 is then dependent from this QS forms the correction amount QD. Of course, in connection with this third realization possibility the difference formation, and thus the signal QS is also replaced by the signal QS, analogously to the first implementation option will.

The fourth possibility for generating the correction quantity QD is that with the aid of the transformation 50 whose input signal 11 = QS is changed so that the Output signal 12 the difference signal of the actual Control path, i.e. 12 = aeist. The replica of the signal aeist can be implemented, for example, with the aid of a delay element of at least the first order. The particular advantage of this fourth implementation option is that an existing in the control Signal, namely QS can be used for the calculation, and still due to the simulation of the Signal aeist the maximum accuracy of the correction amount QD is achieved.

FIG. 8 shows a time diagram for the third and fourth possible implementation of the correction 27 FIG. 8a shows the times of injection and the duration of the injection of fuel with the aid of the metering valve in the pump element, but without correction, the figure 8b actually injected into the internal combustion engine Fuel quantity, again without correction, FIG. 8c shows the injection times and injection periods of fuel in the pump element, however

with correction, and finally FIG. 8d shows the injection times and the injection durations of those actually injected into the internal combustion engine Fuel quantity QK, also again with correction. The individual injection times or injection periods are shown in FIG. 8 with arrows, the tips of the arrows indicating the start of metering or the metering end of the supply of fuel either into the pump element or into the combustion chamber of the internal combustion engine represent. Finally, a load jump is provided in FIG. 8 to identify the correction, where the designation ti an injection before this load change, the designation t2 an injection after identifies this load change.

FIG. 8a shows the injection times and the injection durations the supply of fuel to the pump elements belonging to the individual cylinders injection pump. As can be seen from FIG. 8a, the load jump causes the two injections ti and t2 a shift in the start of metering of fuel into the pump elements at an earlier, undesirable point in time. This is caused by the fact that Due to the sudden change in load, a larger amount of fuel is immediately metered into the pump element without doing so make a correction. Since not only the amount of fuel is changed as a result of the load jump, but also the end of delivery is adapted to the new amount of fuel to be injected, is due to the change At the end of delivery, in this case the base quantity in the pump element is smaller at a later point in time, which causes a temporary increase in the amount of fuel actually injected into the internal combustion engine. These processes have already been explained in more detail in connection with FIG. The said increase which is actually used in the internal combustion engine

injected fuel amount QK due to the decrease the base quantity QS due to the adjustment of the delivery end FE at a later point in time, this undesired The amount of fuel is identified in FIG. 8b with the designation QV. In the representation of the FIG. 8b accordingly too much fuel is injected into the internal combustion engine, which can be attributed to this is that there is no correction. FIG. 8c shows the injection times and the injection durations for the Metering of fuel into the pump element, but this time with a correction of this metered quantity. In FIG. 8c is therefore the amount of fuel actually metered into the pump element with the aid of the metering valve denoted by QE, while the amount of fuel withdrawn in comparison to FIG. 8a is denoted by QD is. The fuel quantity QD is the correction quantity already mentioned several times QD, which can be generated e.g. with the help of the correction options already described. Last shows FIG. 8d shows the injection times and the injection durations actually injected into the internal combustion engine Fuel quantity QK with correction of the fuel quantities metered to the pump element by the metering valve. The figure is therefore analogous to how FIG. 8b had to be viewed in connection with FIG. 8a 8d can be assigned directly to FIG. 8c. From Figure 8d it can now be seen that due to the shown in Figure 8c Correction amount QD does not temporarily affect the amount of fuel actually injected into the internal combustion engine is excessive, as was the case in Figure 8b, but that in Figure 8d, the amount of fuel QK first makes a small jump immediately after the load change, and then slowly to the desired one, the internal combustion engine To increase the amount of fuel to be injected

Overall, with the help of the correction quantity QD, it is possible in the event of load jumps and associated end-of-delivery changes an increase in the amount of fuel actually injected into the internal combustion engine and to avoid a too early start of metering or delivery and at the same time to influence the injection in such a way that that the amount of fuel actually injected into the internal combustion engine slowly increases to the value of the desired amount of fuel to be injected into the internal combustion engine increases.

FIG. 9 shows time diagrams for the fourth possibility of realizing the correction 27. In FIG on the one hand, the simulation of the signal 11 represented by the signal 12, and on the other hand that according to the figure signal QD generated as a function of signal 12. All signals are plotted over the time axis, where this time axis is characterized by the sequence of the cylinders of the internal combustion engine.

According to the explanations relating to FIG. 7, the fourth possibility of realizing the correction 27 is in the case of the signal 11 about the signal relating to the base quantity QS, and in the case of the signal 12 about a replica of the Signal relating to the actual setting angle of the pump piston aeist. This connection, especially the The simulation of the signal aeist by the signal 12 can be seen in the upper diagram in FIG. 9a. How also from the description of the fourth possibility of implementing the correction 27, the signal is then generated with respect to the correction quantity QD as a function of the differences in the simulated signal 12, whereby the difference is always obtained from two values that belong to points in time that one cylinder revolution be apart. The signal QD is plotted as an example in the lower of the two diagrams in FIG. 9a. the

FIG. 9b also shows the signal 12 in its top diagram, but on a different scale, as was the case in FIG. 9a. Next in the Figure 9b also the reference signals and segment signals are shown, as has already been explained k in connection with FIG. Finally, FIG. 9b also shows, by way of example, signal MV1, which, as has already been explained in connection with FIG. K , is used to control the solenoid valve belonging to cylinder 1. In the illustration of this signal MV1, the reference numeral 71 denotes a delay time, the reference numeral 72 denotes a quantity pulse time without correction, while the reference numeral 73 denotes a quantity pulse time lengthening due to the correction and the reference numeral 7 ^ a quantity pulse time shortening also due to the correction .

As has already been stated, a correction quantity QD results, for example according to FIG. which can be positive or negative, i.e. lengthen (73) or shorten (7 * 0 can. If the signal 12 increases in FIG. 9b, this has the consequence that the actual quantity pulse time 72 is lengthened due to the correction, while a decrease in signal 12 causes a shortening (7 * 0 the actual quantity pulse time 72 due to the correction entails.

Overall, with the aid of the fourth implementation option of the correction 27, it is therefore possible to detect deviations in the actually injected fuel amount of the desired To avoid the amount of fuel to be injected. This is, as shown in Figure 9, by the Simulation of the actual setting angle of the pump piston and the lengthening or shortening that depends on it the actual quantity pulse time is reached.

In connection with the figures 1 to 9 described so far, it is now possible to expand the method according to the invention or the device according to the invention for controlling and / or regulating the fuel metering in an internal combustion engine by further functions the associated adjustment of the end of delivery at later points in time due to the injected base quantity, it is particularly advantageous to switch off the supply voltage of the adjustment mechanism 23 for the pump piston lh with a delay compared to the output stage 25 for the metering valve 10 of FIG can be carried out with the help of a corresponding relay, which, for example, can also be made dependent on the speed of the internal combustion engine to correct voltage of the motor vehicle and / or depending on the temperature of the fuel. Both corrections can be implemented, for example, with the aid of at least two-dimensional characteristic maps, with further operating parameters of the internal combustion engine such as the load on the internal combustion engine and / or the speed of the internal combustion engine also being able to be taken into account. If, within the control of FIG. 2, deviations occur over a longer, predeterminable period of time that continuously exceed a predeterminable limit value, it is possible in a particularly advantageous manner to ensure that the device continues to function reliably by increasing the amount of fuel to be metered to the internal combustion engine is reduced. Such a quantity reduction can also be used as a further safety device

tion become active at speeds that exceed certain, predeterminable limit speeds, so that the To keep the speed of the internal combustion engine below the limit mentioned. Similarly, it is also possible e.g. with the help of a speed-dependent limitation map and a subsequent minimum selection level, the driving speed of the motor vehicle driven by the internal combustion engine below to keep from certain, predefinable thresholds. Overall, it is therefore not possible with the aid of the invention only the desired amount of fuel to be supplied to the internal combustion engine to the combustion chambers at the right moment the internal combustion engine, special it is also possible in a particularly advantageous manner to this Injection of fuel into the internal combustion engine targeted monitoring using simple safety devices. The realization of the entire invention is As already mentioned at the beginning, in a particularly advantageous manner with the help of an electronic computing device, so z, B. possible with the help of a suitably programmed microprocessor. With that in connection with this implementation ensures that the entire control system works reliably, it is of great advantage to distribute the individual tasks and functions of the control to two microprocessors coupled to one another. The distribution of the task, as well as the coupling of the processors, is in different ways and ways are possible but do not pose fundamental problems to one skilled in the art. The implementation of the described FIGS. 1 to 9 are also without further inventive steps for a person skilled in the art Activities possible. The selection of the most advantageous alternatives for the respective case, such as the The choice of whether permanently programmed counters or hardwired counters are used depends on the respective Application and the specified boundary conditions.

This is achieved in that the signals for controlling the metering valve and the pump piston are dependent on the operating parameters of the internal combustion engine are corrected.

It is particularly advantageous here to correct the actual or desired operating state of the internal combustion engine to make dependent. Another particularly advantageous development of the invention consists in to simulate the actual operating state of the pump piston, e.g. with the help of a time delay of the signal relating to the desired operating state of the pump piston.

Claims (8)

j 3 5θ 5 4 8 R. j -Λ- " 11/13/198 U Sr / Hm ROBERT BOSCH GMBH, TOOO STUTTGART 1 Expectations mJ A method for controlling and / or regulating the fuel metering in an internal combustion engine with a fuel injection pump which, at least as a function of an electrically controllable metering valve and an electrically adjustable pump piston, injects fuel into the cylinder of the internal combustion engine, as well as with a feedback memory that ψ with the fuel injection pump is in operative connection, characterized denotes overall \ that the actuation of the metering valve and / or * is corrected ■ the pump piston. 2. The method according to claim 1, characterized in that the correction as a function of operating parameters the internal combustion engine is carried out. 3. The method according to claim 2, characterized in that the correction is dependent on the actual operating state of the pump piston. H. Method according to Claim 2, characterized in that the correction is dependent on the desired operating state of the pump piston. 5. The method according to claim 3 or k, characterized in that the correction is dependent on the mean change in the actual or the desired operating state of the pump piston or on the change ORJGINAL INSPECTED -4 - P - '·.' .. · ■ change of the actual or desired operating state of the pump piston compared to a certain, predeterminable previous point in time. 6. The method according to claim 3, characterized in that the actual operating state of the pump piston is simulated. 7. The method according to claim 6, characterized in that the simulation of the actual operating state of the pump piston consists of a time delay in the desired operating state of the pump piston. 8. Device for carrying out the method for controlling and / or regulating the fuel metering in an internal combustion engine according to at least one of claims 1 to 7, with a fuel injection pump, at least depending on one electrically controllable metering valve and one electrically adjustable Pump piston performs the injection of fuel into the cylinder of the internal combustion engine, as well as with a return memory which is in operative connection with the fuel injection pump, thereby characterized in that at least one electronic computing device is present, with the aid of which the metering valve and the pump piston can be controlled.