Method for operating an internal combustion engine with gaseous fuel, computer program product and control device

Abstract

The invention relates to a method for operating an internal combustion engine (10), in which method gaseous fuel is supplied from a fuel rail (18) and via an injector (20) assigned to a respective cylinder (14) by injection, there is a pause between injection for one cylinder (14) and subsequent injection for another cylinder (14) in which no injection takes place, and gaseous fuel is supplied to the fuel rail (18) substantially continuously from a fuel supply system (16). It is proposed that the method comprises the following steps: (a) determining a first variable, which characterises the inflow of gaseous fuel into the fuel rail (18); (b) determining a second variable, which at least indirectly characterises the sum of the quantity of fuel removed from the fuel rail (18) during injection and the quantity of fuel supplied to the fuel rail (18) by the fuel supply system (16) during injection; (c) determining a third variable, which characterises the quantity of fuel removed from the fuel rail (18) due to the injection, using the first and the second variable.

Description

[0001] R.415517

[0002] - 1 -

[0003] Description

[0004] title

[0005] Method for operating an internal combustion engine with gaseous fuel, computer program product and control unit

[0006] State of the art

[0007] The invention relates to a method for operating an internal combustion engine with gaseous fuel, a computer program product and a control unit.

[0008] German patent DE 10 2021 210 001 A1 describes a method for operating an internal combustion engine that runs on gaseous fuel, for example, hydrogen. Air is supplied to a plurality of combustion chambers. Gaseous hydrogen is injected directly into the combustion chambers, or it is injected into the intake ports (port fuel injection), and the hydrogen-air mixture is ignited in the combustion chambers.

[0009] Disclosure of the invention

[0010] The problem underlying the invention is solved by a method with the features of claim 1, as well as by a computer program product and a control unit with the features of the dependent claims. Advantageous embodiments are specified in the dependent claims.

[0011] An advantage of the invention is that, despite the continuous supply of gaseous fuel to the fuel rail, the actual amount of fuel drawn from the fuel rail during an injection, or at least a characteristic parameter thereof, can be detected. This allows, for example, the diagnosis and, if necessary, correction of an injector malfunction. This is achieved through process steps that are easy to implement in software, without the need for additional sensors. R.415517

[0012] - 2 -

[0013] Specifically, this is achieved through a method for operating an internal combustion engine in which gaseous fuel is supplied from a fuel rail and injected into a cylinder via an injector at the desired time. The internal combustion engine can be constructed very similarly to conventional piston engines that use liquid fuel such as diesel or gasoline. The gaseous fuel can be, for example, hydrogen. The injector can inject the gaseous fuel directly into its assigned combustion chamber or cylinder, or it can inject the gaseous fuel into the intake port associated with the combustion chamber or cylinder. In both cases, the injectors are fluidically connected to a common fuel rail, which supplies them with gaseous fuel at a predetermined pressure level.The fuel rail can be a typical, for example tubular, fuel collection line, as is known from gasoline and diesel internal combustion engines with direct injection.

[0014] In the method according to the invention, between an injection for one cylinder and a subsequent injection for another cylinder, there is at least a temporary pause, i.e., a period during which no injection takes place. This pause need not always be present. However, it can be present at at least one "typical" operating point of the internal combustion engine so that the behavior of the injectors can be determined at that operating point. The behavior of the injectors determined at this operating point is then also applied to other operating points. The withdrawal of gaseous fuel from the fuel rail thus occurs discontinuously with withdrawal phases and intervening withdrawal pauses. In contrast, gaseous fuel is supplied to the fuel rail essentially continuously from a fuel supply system. The gaseous fuel is therefore supplied both during the withdrawal phases and during the withdrawal pauses.It is understood that the amount of fuel supplied can be varied depending on the amount extracted by the injectors. The method according to the invention is particularly preferred when the internal combustion engine is in a more or less stationary or at least not too dynamic operating state, i.e., when there is no load and R.415517.

[0015] - 3 -

[0016] The rotational speed does not change, or at least not significantly, and thus the amount of fuel supplied to the fuel rail and the amount of fuel extracted from the fuel rail by the injectors do not change, or at least not significantly.

[0017] According to the invention, the method comprises the following steps:

[0018] (a) Determining a first quantity that characterizes the inflow of gaseous fuel into the fuel rail. This first quantity thus provides knowledge of the quantity of fuel continuously supplied to the fuel rail per unit of time. This quantity is typically the quantity supplied to the fuel rail by the fuel supply system. The fuel supply system typically comprises a fuel reservoir in which the fuel is stored under high pressure. By means of a suitable device, the fuel is supplied to the fuel rail at a desired and comparatively low pressure via several pressure regulators.

[0019] (b) Determining a second quantity that at least indirectly characterizes the sum of, on the one hand, the quantity of fuel withdrawn from the fuel rail during an injection and, on the other hand, the quantity of fuel supplied to the fuel rail by the fuel supply system during that same injection. This second quantity thus characterizes a kind of mass balance of the fuel rail, i.e., a balance of what gaseous fuel is withdrawn from the fuel rail during an injection by an injector and what gaseous fuel is simultaneously supplied to the fuel rail by the fuel supply system during that injection.

[0020] The following consideration underlies this: In contrast to injection systems with liquid fuels (e.g., gasoline, diesel), the pressure level in a fuel rail with gaseous fuel is not regulated by a discontinuously supplying pump, such as a piston pump, but by the continuous supply mentioned above, provided by a suitable pressure regulator supplied from the aforementioned fuel tank. The supply is at least approximately constant and should, over a complete operating cycle of the combustion engine, be at least approximately equal to the sum of all fuel drawn via the injectors.

[0021] - 4 -

[0022] Gas quantities correspond. With one injection per cylinder and operating cycle, the resulting rail pressure curve ideally resembles a regular sawtooth pattern. The rising edges result from the continuous and constant supply (first parameter), while the falling edges result from the superposition of gas withdrawal and parallel supply, i.e., the second parameter.

[0023] (c) Determining a third quantity, which characterizes the amount of fuel extracted from the fuel rail by injection, using the first and second quantities. Typically, the determination of the third quantity may also involve calculating the difference between the first and second quantities.

[0024] In a further training course, it is stipulated that step (c) be carried out for each cylinder of the internal combustion engine, and that the procedure additionally includes the following step (d): correcting the injector control so that the third parameters have at least approximately the same value. This is based on the following considerations: Due to manufacturing variations in injector production and changes over time (e.g., wear), the injectors do not meter the fuel exactly the same. Without countermeasures, a deviation in metering across the cylinders of the internal combustion engine is unavoidable, resulting from differences in switching behavior and / or steady-state flow. The differences in switching behavior manifest themselves, for example, in different opening and / or closing times.Differences in steady-state flow can be caused, for example, by different opening widths of the injector outlets. The continuous supply of fuel in the present gaseous fuel system is a significant difference compared to gasoline or diesel injection systems. This prevents the injector control from being equalized using previously known approaches. The correction now proposed in step (d), together with steps (a) - (c), makes it possible to equalize the fuel delivery of the different injectors even in a gaseous fuel system. This improves the smooth running of the internal combustion engine. R.415517.

[0025] - 5 -

[0026] In other words, without correcting the injector control, the control for all injectors would be calculated based on predefined target quantities and an assumed "nominal behavior" of the injectors (flow rate, switching dynamics). The sequential extraction of fuel from the fuel rail by the injectors, combined with the parallel and continuous replenishment of fuel from the fuel supply system into the fuel rail, creates a sawtooth pressure profile in the fuel rail. This profile is measured, and indicators for all injectors are calculated from it, as described above—namely, the third parameter described above. The relative comparison of these indicators provides information about which injector is extracting more or less fuel than the average across all injectors. Based on this information, individual corrections are created for each injector.These corrections were adjusted to ensure that the injectors extract or inject at least approximately the same amount of fuel.

[0027] If the differences in injected quantities are due to flow variations in the injectors (so-called "Qstat differences"), a multiplicative correction makes sense. If the injection differences are caused by different opening and closing times, an additive correction is more advantageous. The corrections can be continuously adjusted, creating a closed control loop that reduces the deviations. Alternatively, it is also possible to first observe the indicators over a longer period and then subsequently calculate and apply a correction averaged over the observation period, for example, when restarting the internal combustion engine.

[0028] In a further training course, the first parameter is intended to characterize a pressure gradient in the fuel rail during a pause between two successive injections. This parameter can be easily determined, either by measurement using a sensor or by a software model.

[0029] In a further training course, it is stipulated that the first quantity is determined either using the pressure and time at the end of an injection and the pressure and time at the beginning of a subsequent injection, or that the first quantity is determined as the mean value of pressure gradients between R.415517

[0030] - 6 - the first value is determined at the end of one injection and the beginning of a subsequent injection, or that the first value is determined as the average pressure gradient during a partial period between the end of one injection and the beginning of a subsequent injection, or that the first value is determined using the maximum pressure gradient in the period between the end of one injection and the beginning of a subsequent injection. All of the above variants can be easily implemented in software.

[0031] In a further development, the first value is provided for as an average – for example, an arithmetic mean – calculated from values ​​obtained during different intervals between two injections. In this way, statistical outliers can be eliminated, and consequently, the accuracy of the method according to the invention can be further improved.

[0032] In a further training course, the second quantity is to be determined using a pressure in the fuel rail and a corresponding time before injection, and a pressure in the fuel rail and a corresponding time after injection. This involves the difference between the pressure in the fuel rail before injection and the pressure in the fuel rail after injection, and a corresponding time difference; or, in other words, two pairs of values ​​p1|t1 and p2|t2. These quantities can also be determined easily and with high accuracy and resolution.

[0033] The training program stipulates that the first and / or second parameter must also be measured by a sensor. Such a pressure sensor is almost always already present on the fuel rail, so this does not incur any additional costs.

[0034] In a further training course, it is stipulated that the third quantity is a pressure difference, calculated as the sum of the product of a time period and the pressure gradient, and the pressure difference in the fuel rail between the beginning and end of that time period. This is calculated using the aforementioned pressure gradient dp / dt (first quantity) and, as the second quantity, the two values ​​R.415517.

[0035] - 7 -

[0036] From the pairs of values ​​p1|t1 and p2|t2, which at least indirectly characterize the sum of the amount of fuel extracted from the fuel rail during injection and the amount of fuel supplied to the fuel rail by the fuel supply system during this time, a pressure difference Ap = p1 - p2 + dp / dt ■ (t2 - 11) can be calculated as a third quantity, in which the continuous supply of gaseous fuel to the fuel rail is "calculated out" and which therefore correlates with the amount of gaseous fuel injected by the injector.

[0037] Examples of the value pairs p1|t1 and p2|t2 could be: pressure maximum before and pressure minimum after injection; pressure maximum before and pressure maximum after injection; pressure values ​​in the middle of the rising flank before and after injection (i.e., temporally, on the one hand, in the middle between the end of the previous injection and the currently considered injection, and on the other hand, in the middle between the end of the current injection and the beginning of the subsequent injection); pressure values ​​at the time of the maximum pressure gradient in the rising flanks before and after injection.

[0038] The invention also relates to a computer program product comprising instructions which, when the program is executed by a computer, cause it to execute a method of the type described above.

[0039] The invention also relates to a control unit for controlling and / or regulating the operation of an internal combustion engine, comprising at least one processor, at least one memory, and at least one computer program product of the type just described, stored in the memory. The control unit can also be referred to as a computer.

[0040] The following are an explanation of embodiments of the invention with reference to the drawing. The drawing shows:

[0041] Figure 1 shows a schematic representation of a six-cylinder internal combustion engine in which gaseous fuel is injected directly into the respective cylinders from a fuel rail via injectors; R.415517

[0042] - 8 -

[0043] Figure 2 is a diagram showing the pressure profile in the fuel rail of Figure 1 over time for injections into all six cylinders of Figure 1;

[0044] Figures 3-7 are diagrams similar to Figure 2, but only for two injections, to illustrate different methods for determining a first quantity in the form of a pressure gradient dt / dt;

[0045] Figures 8-11 are diagrams similar to Figures 3-7 to illustrate different methods for determining a second quantity in the form of pressure / time value pairs;

[0046] Figure 12 is a bar chart in which a pressure difference in the fuel rail of Figure 1 due to the injection of fuel into the cylinders of Figure 1 is plotted as a third quantity, without correction of the control of the injectors;

[0047] Figure 13 is a bar chart similar to Figure 12, but with correction of the injector control;

[0048] Figure 14 shows a diagram in which the pressure in the fuel rail of an alternative four-cylinder internal combustion engine is plotted over time, without correction of the injector control;

[0049] Figure 15 shows a diagram similar to Figure 14, but with a correction to the injector control; and

[0050] Figure 16 shows a flowchart of a method for operating the internal combustion engine of Figure 1, with correction of the control of the injectors.

[0051] In Figure 1, an internal combustion engine is designated by reference numeral 10. It comprises an engine block 12 with, by way of example, six cylinders 14a-f. Furthermore, the internal combustion engine 10 comprises a fuel supply system 16, which supplies gaseous fuel to a fuel rail 18. (Reference to R.415517)

[0052] - 9 -

[0053] In the present example, six injectors 20a-f are connected to fuel rail 18. Each injector 20a-f is assigned to a cylinder 14a-f. This allows injector 20a to inject the gaseous fuel individually for cylinder 14a, injector 20b to inject the gaseous fuel individually for cylinder 14b, and so on. This can be achieved by the injectors 20a-f injecting the gaseous fuel directly into their respective assigned cylinders 14a-f, as shown in the present example. In an embodiment not shown, this can be achieved by the injectors injecting the gaseous fuel into the intake manifold of each respective cylinder.

[0054] Each injector 20a-f has an actuator (not shown), which could be, for example, an electromagnetic or a piezoelectric actuator. This actuator allows an (unshown) and often needle-like valve element to be switched from a closed position to an open position and back again for injection. Each injector 20a-f is controlled separately for this purpose, as indicated by the respective dashed arrows.

[0055] A control unit 22, which includes at least one processor 24, at least one memory 26, and at least one computer program product 28 stored in the memory 26, serves to control, among other things, the injectors 20a-f. The control unit 22 can also be referred to as a computer or comprise one. The computer program product 28 comprises instructions which, when the program is executed on the processor 24, cause the control unit 22 to execute certain process steps corresponding to the computer program product 28, as will be described below. The internal combustion engine 10 also includes a pressure sensor 30, which detects the pressure of the gaseous fuel in the fuel rail 18 with high temporal resolution and delivers a corresponding signal to the control unit 22.

[0056] Due to manufacturing variations in the production of injectors 20a-f and changes over their service life (e.g., wear), the injectors 20a-f do not meter the fuel exactly the same – unless countermeasures are taken. Without appropriate countermeasures, this inevitably leads to R.415517.

[0057] - 10 - a deviation in the metering via cylinders 14a-f of the internal combustion engine 10, resulting from differences in switching behavior and / or steady-state flow. The differences in switching behavior manifest themselves, for example, in different opening and / or closing times, for instance, due to manufacturing tolerances of the actuators. The differences in steady-state flow can be caused, for example, by different opening widths of the injector outlet ports. All of this can lead to rough engine running, undesirable pollutant emissions, especially NOx, reduced efficiency with increased fuel consumption, and even engine damage, for example, due to knocking combustion.It is therefore desirable to correct the control of the injectors 20a-f by the control unit 22 so that, at least during steady-state operation of the internal combustion engine 10, the same quantity of fuel is injected into each cylinder 14a-f sequentially during a single operating cycle in which fuel (possibly divided into several partial injections) is injected into each cylinder 14a-f. A method for such a correction is now explained.

[0058] The sequential extraction of fuel from the fuel rail 18 by the injectors 20a-f, combined with the parallel and continuous supply of fuel from the fuel supply system 16 into the fuel rail 18, results in a sawtooth pressure profile in the fuel rail 18. This profile is measured by the pressure sensor 30. The uncorrected pressure profile is shown in Figure 2. Falling flanks are designated 32a-f, and rising flanks are designated 34a-f. Falling flanks 32a-f are generated by the respective injections into the cylinders 14a-f and the simultaneous continuous supply of fuel from the fuel supply system 16.Rising flanks 34a-f indicate the pauses after each injection into a cylinder 14a-f, during which no fuel is taken from the fuel rail 18 via the injectors 20a-f and instead fuel is continuously supplied from the fuel supply system 16 into the fuel rail 18.

[0059] These rising flanks 34a-f are used to determine a first quantity that characterizes the inflow of gaseous fuel into the fuel rail 18. This first quantity is, for example, R.415517.

[0060] - 11 -

[0061] Pressure gradient dp / dt, namely the slope of a rising flank 34a-f. In Figures 3-7, various possibilities are shown for determining this pressure gradient dp / dt for the rising flank 34a after the injection 32a into the cylinder 14a and before the injection 32b into the cylinder 14b.

[0062] According to Figure 3, the first parameter dp / dt is determined using the pressure p(min) and time t(min) at the end of an injection 32a and the pressure p(max) and time t(max) at the beginning of a subsequent injection 32b by simply subtracting (p(max) - p(min)) / (t(max) - t(min)). According to Figure 4, the first parameter dp / dt is determined as the mean of a plurality of point-by-point measured pressure gradients dp / dt between the end of injection 32a and the beginning of the subsequent injection 32b, for example, based on the measured pressure signal via a linear fit or a digital filter (HR, FIR, or a combination thereof). The input values ​​are then the raw pressure values, and the output values ​​are the pressure gradient.

[0063] As shown in Figures 5-7, the first quantity dp / dt is determined as the mean (for example, as the arithmetic mean) of local pressure gradients during a partial period between the end of injection 32a and the beginning of the subsequent injection 32b. Specifically, according to Figure 5, this occurs shortly before the maximum pressure or the beginning of the subsequent injection 32b; according to Figure 6, during a partial period approximately midway between the two injections 34a and 34b; and according to Figure 7, at the point during the pause or rising flank 34a when the rising flank 34a has its maximum slope. It is also possible to determine the first quantity or pressure gradient dp / dt during several different pauses or rising flanks, for example, during several consecutive rising flanks 34a-f, and then calculate a mean value, such as an arithmetic mean.

[0064] Furthermore, a second quantity is determined, which at least indirectly characterizes the sum of, on the one hand, the amount of fuel extracted from the fuel rail 18 during an injection and, on the other hand, the amount of fuel supplied to the fuel rail 18 by the fuel supply system 16 during this time. Specifically, in the present embodiment, the second quantity is R.415517

[0065] - 12 -

[0066] The size is determined using a pressure p1 and a corresponding time t1 before an injection 32b and a pressure p2 and a corresponding time t2 after the injection 32b. In this way, two pairs of values ​​p1|t1 and p2|t2 are generated. There are also various ways to determine these two pairs of values ​​p1|t1 and p2|t2, which are illustrated by way of example in Figures 8-11:

[0067] According to Figure 8, the p1|t1 value pair is determined at the maximum pressure immediately before injection 32b, and the p2|t2 value pair is determined at the minimum pressure immediately after injection 32b. According to Figure 9, the p1|t1 value pair is determined at the maximum pressure immediately before injection 32b, and the p2|t2 value pair is determined at the (first) maximum pressure after injection 32b. According to Figure 10, the p1|t1 value pair is determined in the middle of the rising flank 34a before injection 32b, and the p2|t2 value pair is determined in the middle of the rising flank 34b after injection 32b. According to Figure 11, the value pair p1 |t1 is determined at the time of the maximum pressure gradient dp / dt of the rising flank 34a before the injection 32b, and the value pair p2|t2 is determined at the time of the maximum pressure gradient dp / dt of the rising flank 34b after the injection 32b.Optionally, filtering can be used as a preprocessing step to smooth and reduce measurement noise.

[0068] The actual pressure difference in the fuel rail 18 solely due to the amount of fuel injected by the injector 20a during injection 32b into the cylinder 14b, i.e. the “net pressure difference”, from which the pressure increase due to the continuous supply of fuel from the fuel supply system 16 into the fuel rail 18 is calculated, is then given by the following formula:

[0069] Ap = p1 - p2 + dp / dt ■ (t2 - 11)

[0070] This pressure difference Ap is determined for all cylinders 14a-f for at least one complete operating cycle. If necessary, the pressure differences Ap are also determined for several complete operating cycles for each cylinder 14a-f and then averaged for each cylinder 14a-f, for example as an arithmetic mean. The determined pressure differences Ap form indicators that provide an R.415517

[0071] - 13 -

[0072] This provides information on which injector 20a-f extracts more or less fuel than the average across all injectors 20a-f. The corresponding indicators are shown as bar charts in Figure 12, where the indicators bear the reference numeral 36 and are labelled with the previously known indices af for cylinders 14a-f.

[0073] For example, the arithmetic mean of indicators 36a-f can now be determined, and then the deviations of the individual indicators 36a-f from the arithmetic mean can be calculated. Based on these deviations, correction values ​​for the control of the injectors 20a-f by the control unit 22 can now be determined, resulting in all indicators 36a-f having the same value, as shown in Figure 13. This means that all injectors 20a-f supply the same amount of fuel to the corresponding cylinder 14a-f during a work cycle under steady-state operating conditions of the internal combustion engine 10.

[0074] If the differences in the injected quantities are due to flow deviations of the injectors 20a-f (so-called "Qstat differences"), a multiplicative correction makes sense. If the injection differences are caused by different opening and closing times, an additive correction is more advantageous. The corrections can be continuously adjusted, creating a closed control loop that reduces the deviations. Alternatively, it is also possible to first observe the indicators 36a-f over a longer period and then subsequently calculate and apply a correction averaged over the observation period, for example, when restarting the internal combustion engine 10.

[0075] Figure 14 shows a pressure / time diagram of the pressure in a fuel rail of a four-cylinder internal combustion engine 10 before a correction, and Figure 15 shows a pressure / time diagram of the pressure in a fuel rail of a four-cylinder internal combustion engine 10 after such a correction.

[0076] For the sake of completeness, the procedure described above for determining indicators 36a-f is now also explained with reference to the flowchart in Figure 16: R.415517

[0077] - 14 -

[0078] In a function block 38, the target quantities to be injected for each injector 20a-f during an injection cycle are specified or determined, for example, taking into account the rotational speed of the internal combustion engine 10 and a load requirement from a user, i.e., taking into account a desired torque. Based on this, among other things, the control signals for the injectors 20a-f are determined in a function block 40 based on their nominal behavior, i.e., an ideal flow rate and ideal switching dynamics. The injectors 20a-f are initially controlled with these control signals. The pressure in the fuel rail 18 is measured with high precision and high temporal resolution using the pressure sensor 30. In a function block 42, the corresponding indicators 36a-f are determined for each injector 20a-f according to the procedure described above.In function block 44, individual correction values ​​are determined for each injector 20a-f, with the aim that the indicators 36a-f all have the same value. The correction values ​​determined in function block 44 are then also fed into function block 40, so that the injectors 20a-f are controlled with correspondingly adjusted control signals.

Claims

R.415517 - 15 - Claims 1. A method for operating an internal combustion engine (10) in which gaseous fuel is supplied from a fuel rail (18) and via an injector (20) assigned to a respective cylinder (14) by means of an injection (32), in which at least temporarily between an injection (32) for one cylinder (14) and a subsequent injection (32) for another cylinder (14) there is a pause (34) in which no injection (32) takes place, and in which gaseous fuel is supplied to the fuel rail (18) substantially continuously from a fuel supply system (16), characterized in that it comprises the following steps: (a) Determining a first quantity which characterizes the inflow of gaseous fuel into the fuel rail (18).; (b) Determining a second quantity which at least indirectly characterizes the sum of, on the one hand, the quantity of fuel withdrawn from the fuel rail (18) during an injection (32) and, on the other hand, the quantity of fuel supplied to the fuel rail (18) during that same injection by the fuel supply system (16); (c) Determining a third quantity which characterizes the quantity of fuel withdrawn from the fuel rail (18) by the injection (32) using the first and the second quantity.

2. The method of claim 1, characterized in that step (c) is carried out for each cylinder (14) of the internal combustion engine (10), and that the method additionally comprises the following step: (d) correcting the control of the injectors (20) so that the third quantities have at least approximately the same value.

3. Method according to at least one of claims 1-2, characterized in that the first quantity characterizes a pressure gradient in the fuel rail (18) in a pause (34) between two successive injections (32). R.415517 - 16 - 4. Method according to claim 3, characterized in that the first quantity is determined either using the pressure and time at the end of an injection (32) and the pressure and time at the beginning of a subsequent injection (32), or that the first quantity is determined as the mean value of pressure gradients between the end of an injection (32) and the beginning of a subsequent injection (32), or that the first quantity is determined as the mean value of pressure gradients during a partial period between the end of an injection (32) and the beginning of a subsequent injection (32), or that the first quantity is determined using the maximum pressure gradient in the period between the end of an injection (32) and the beginning of a subsequent injection (32).

5. Method according to at least one of claims 3-4, characterized in that the first quantity is an average value formed from values ​​determined during different pauses (34) between two injections (32).

6. Method according to at least one of the preceding claims, characterized in that the second quantity is determined using a pressure and an associated time before an injection (32) and a pressure and an associated time after the injection (32).

7. Method according to claim 6, characterized in that the third quantity is a pressure difference and is determined as the sum of, on the one hand, the product of a time period with the pressure gradient and, on the other hand, the pressure difference in the fuel rail (18) between the beginning and end of the time period.

8. Computer program product (28), comprising instructions which, when the program is executed by a computer (22), cause the computer to execute the method according to at least one of the preceding claims.

9. Control unit (22) for controlling and / or regulating the operation of an internal combustion engine (10), comprising at least one processor (24), R.415517 - 17 - at least one memory (26) and at least one computer program product (28) stored on the memory according to claim 8.

Images