METHOD FOR ASCERTAINING A VARIABLE CHARACTERIZING A FLOW RATE OF A FUEL INJECTOR

Abstract

A method for ascertaining a variable characterizing a flow rate of a fuel injector during an operation of an internal combustion engine, to which the fuel injector is assigned. At least two input values for a data-based model are ascertained, and at least one output value is determined with the aid of the data-based model, on the basis of which a value for the variable characterizing the flow rate of the fuel injector is ascertained. The data-based model combines at least two methods differing from one another for ascertaining a variable characterizing a flow rate of a fuel injector.

Claims

1-15. (canceled)

16. A method for ascertaining a variable characterizing a flow rate of a fuel injector during an operation of an internal combustion engine to which the fuel injector is assigned, the method comprising the following steps: ascertaining at least two input values for a data-based model; determining at least one output value with the aid of the data-based model; and ascertaining, based on the at least one output value, a value for the variable characterizing the flow rate of the fuel injector; wherein the data-based model combines at least two methods differing from one another for ascertaining the variable characterizing the flow rate of the fuel injector.

17. The method as recited in claim 16, wherein the at least one output value is ascertained with the aid of the data-based model, using machine learning.

18. The method as recited in claim 17, wherein the data-based model is an artificial neural network.

19. The method as recited in claim 16, wherein the variable characterizing the flow rate of the fuel injector includes an absolute or relative flow rate of the fuel injector.

20. The method as recited in claim 16, wherein at least one activation parameter for the fuel injector is adapted based on the ascertained value for the variable characterizing the flow rate of the fuel injector.

21. The method as recited in claim 16, wherein the data-based model includes a method for ascertaining a variable characterizing a flow rate of a fuel injector, in which a pressure-dependent measure of the variable is determined based on a return point in time of a nozzle needle of the fuel injector and/or on a closing duration of the nozzle needle of the fuel injector.

22. The method as recited in claim 16, wherein the data-based model includes a method for ascertaining a variable characterizing a flow rate of a fuel injector, in which a measure of the variable is determined in a range of a closing point in time of a nozzle needle of the fuel injector based on characteristics of a profile of a signal of a sensor, which is provided for detecting an opening and/or closing of the fuel injector.

23. The method as recited in claim 16, wherein the data-based model includes a method for ascertaining a variable characterizing a flow rate of a fuel injector, in which a measure of an injected quantity of fuel is determined based on a start of delivery of a high-pressure pump to determine a measure of the variable.

24. The method as recited in claim 16, wherein the data-based model includes a method for ascertaining a variable characterizing a flow rate of a fuel injector, in which a measure of the variable is determined based on a pressure change caused by a main injection of the fuel injector, and/or based on a pressure gradient in a high-pressure accumulator via which the fuel injector is supplied.

25. The method as recited in claim 16, wherein the data-based model includes a method for ascertaining a variable characterizing a flow rate of a fuel injector, in which a measure of the variable is determined based on a ratio of a quantity of fuel introduced during an injection of the fuel injector and a time duration of the injection.

26. The method as recited in claim 16, wherein geometric parameters of the fuel injector are taken into consideration in the data-based model.

27. The method as recited in claim 16, wherein at least two input values for the data-based model are selected or derived from values of the following variables: return point in time of a nozzle needle of the fuel injector, closing duration of the nozzle needle of the fuel injector, characteristics of a profile of a signal of a sensor which is provided for detecting an opening and/or closing of the fuel injector in a range of a closing point in time of the nozzle needle of the fuel injector, start of delivery of a high pressure pump, a pressure change in a high pressure accumulator, caused by a main injection of the fuel injector, via which the fuel injector is supplied, pressure gradients in the high pressure accumulator, via which the fuel injector is supplied, caused by a main injection of the fuel injector, a ratio of a quantity of fuel introduced during an injection of the fuel injector and a time duration of the injection.

28. A processing unit configured to ascertain a variable characterizing a flow rate of a fuel injector during an operation of an internal combustion engine to which the fuel injector is assigned, the processing unit configured to: ascertain at least two input values for a data-based model; determine at least one output value with the aid of the data-based model; and ascertain, based on the at least one output value, a value for the variable characterizing the flow rate of the fuel injector; wherein the data-based model combines at least two methods differing from one another for ascertaining the variable characterizing the flow rate of the fuel injector.

29. A non-transitory machine-readable memory medium on which is stored a computer program for ascertaining a variable characterizing a flow rate of a fuel injector during an operation of an internal combustion engine to which the fuel injector is assigned, the computer program, when executed by a processing unit, causing the processing unit to perform the following steps: ascertaining at least two input values for a data-based model; determining at least one output value with the aid of the data-based model; and ascertaining, based on the at least one output value, a value for the variable characterizing the flow rate of the fuel injector; wherein the data-based model combines at least two methods differing from one another for ascertaining the variable characterizing the flow rate of the fuel injector.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0030] FIG. 1 schematically shows an internal combustion engine including a common rail system, which is suitable for carrying out a method according to the present invention.

[0031] FIG. 2 schematically shows a fuel injector, in which a method according to the present invention is implementable.

[0032] FIG. 3 schematically shows a signal profile, on the basis of which a part of a method according to the present invention in one preferred specific embodiment is explained.

[0033] FIG. 4 schematically shows a pressure profile, on the basis of which a part of the method according to the present invention in one preferred specific embodiment is explained.

[0034] FIG. 5 schematically shows model data of a data-based model, which may be used in a method according to the present invention in one preferred specific embodiment.

[0035] FIG. 6 schematically shows a comparison for the accuracy in a method according to the present invention in one preferred specific embodiment and in a method not according to the present invention.

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS

[0036] An internal combustion engine 100 is schematically shown in FIG. 1, which is suitable for carrying out a method according to the present invention. Internal combustion engine 100 includes, for example, three combustion chambers and associated cylinders 105. Each combustion chamber 105 is assigned a fuel injector 130, each of which, in turn, is connected to a high pressure accumulator 120, a so-called rail, via which it is supplied with fuel. It is understood that a method according to the present invention may also be carried out in an internal combustion engine including an arbitrary number of cylinders, for example, four, six, eight or twelve cylinders. It is also understood that a method according to the present invention may also be carried out in an internal combustion engine that includes more than one rail, for example, V-engines including two rails.

[0037] Furthermore, high pressure accumulator 120 is fed fuel from a fuel tank 140 via a high pressure pump 110. High pressure pump 110 is coupled to internal combustion engine 100, namely, in such a way that the high pressure pump is driven via a crankshaft of the internal combustion engine, or via a camshaft which, in turn, is coupled to the crankshaft.

[0038] An activation of fuel injectors 130 for metering or injecting fuel into respective combustion chambers 105 takes place via a processing unit designed as an engine control unit 180. For the sake of clarity, only the connection from engine control unit 180 to one fuel injector 130 is represented; it is understood, however, that each fuel injector 130 is connected accordingly to the engine control unit. Each fuel injector 130 in this case may be specifically activated. Engine control unit 180 is further configured to detect the fuel pressure in high pressure accumulator 120 with the aid of a pressure sensor 190.

[0039] A fuel injector 130 including an actuator 135 designed as a solenoid valve (the procedure may, in principle, also be applied in a fuel injector that includes a piezo actuator) and associated NCS sensor 136, as it may be used, for example, in internal combustion engine 100 according to FIG. 1, is shown by way of example in FIG. 2. Actuator 135 is used to activate a servo valve 134 in fuel injector 130. A nozzle needle 137 is also shown, which is able to unblock and close through-flow openings or nozzle holes 138 for injecting fuel.

[0040] Sensor 136 is situated, for example, at fuel injector 130 in such a way that the sensor signal responds to pressure changes in the valve chamber, as a result of which characteristic points in time of injection processes with the aid of fuel injector 130 such as opening and closing of the nozzle needle or valve needle may then be deduced. Actuator 135 is connected with two activation lines, for example, to an output phase in engine control unit 180. Sensor 136 is connected here via two inputs to engine control unit 180.

[0041] A signal profile S over a time t is schematically shown in FIG. 3, on the basis of which a part of a method according to the present invention in one preferred specific embodiment is explained. Signal profile S represents a profile of the signal supplied by an NCS, which is a voltage, for example.

[0042] On the basis of this profile, it is possible, in principle, to determine a point in time characteristic of the injection process, such as an opening point in time t.sub.o, a return point in time t.sub.u of the nozzle needle as well as a closing point in time t.sub.s. In this way, it is also possible, for example, to determine closing duration Δt.sub.s of the nozzle needle (as i.sub.s-t.sub.u).

[0043] Taking the prevailing pressure in the high pressure accumulator in this case into consideration, it is possible based on these timing variables, in principle, to deduce the through-flow through the fuel injector, and thus a coking. As mentioned, the needle closing duration is, however, not solely a function of the coking of the injector, but also of the guide clearance of the nozzle needle, which is usually constant.

[0044] Against this background, this procedure may also be combined in the data-based model with further, various methods or approaches for ascertaining a flow rate of a fuel injector.

[0045] One further such method is also explained with reference to FIG. 3. The range around closing point in time i.sub.s may be seen in the enlarged detail of signal profile S represented. The specific profile there may be characterized, for example, based on particular characteristics such as signal drop ΔS.sub.1, signal rise ΔS.sub.2 (or also minimum or maximum value with respect to a reference, or also its difference), as well as distance Δt.sub.1 between the minimum value and maximum value in this area.

[0046] This characteristic signal shape is a function of sensor parameters and/or fuel injector parameters such as the nozzle through-flow or the coking. The characteristic variables may be incorporated within the scope of a further approach as parameters into the combined data model.

[0047] A pressure profile having a pressure p over time t is schematically shown in FIG. 4, on the basis of which a part of a method according to the present invention in one preferred specific embodiment is explained. In this case, p.sub.1 indicates the real pressure profile in high pressure accumulator, p.sub.2 in contrast indicates the profile of this pressure ultimately present as a result of measurement in the engine control unit, which is not only discretized, but also time-delayed.

[0048] The determination of the through-flow from gradient dp/dt of a, for example, undisrupted drop of pressure p in the high pressure accumulator during injection takes place, for example, by determining the pressure at two defined points in time t.sub.1 at the start and t.sub.2 at the end of the pressure drop triggered by the injection and by forming the pressure difference. Gradient dp/dt of the pressure drop derived in this manner represents a direct measure of the nozzle through-flow or the coking when the nozzle needle is opened.

[0049] The conditions and tolerances of the detection of the pressure in the high pressure accumulator influence the quality of the model. Thus, the real-time pressure in the high pressure accumulator is generally not available in the engine control unit, as may be seen in the diagram, but merely a filtered time-discrete signal.

[0050] As mentioned, a further procedure for ascertaining the coking is the determination of the ratio of a quantity of fuel introduced during an injection of the fuel injector and a time duration of the injection as a direct measure of the coking. The injection quantity may, as mentioned, be determined via pressure drop Δp between two defined points in time t.sub.0 and t.sub.3 before and after the injection, which is shown in FIG. 4. The injection duration is generally known in the engine control unit or may be ascertained as explained, for example, with reference to FIG. 3 (then there as i.sub.s-t.sub.o). Within the scope of the method, it is provided for data-based model 200 to use individual discrete features of a pressure profile in the high pressure accumulator, i.e., individual discrete measured pressures of a pressure profile in the high pressure accumulator. In this case, discrete features/pressures of a pressure drop of the pressure profile in the high pressure accumulator are to be used. In particular, four, in particular preferably exactly four determined pressures ascertained by measuring are to be used.

[0051] These pressures are, in particular, pressure p.sub.2 at point in time t.sub.0(p.sub.2(t.sub.0)), pressure p.sub.2 at point in time t.sub.1(p.sub.2(t.sub.1)), pressure p.sub.2 at point in time t.sub.0(p.sub.2(t.sub.2) and pressure p.sub.2 at point in time t.sub.3(p.sub.2(t.sub.3)). This means, that pressure p.sub.2 (t.sub.0) is to be used—in particular immediately—prior to the injection, pressure p.sub.2(t.sub.3)—in particular immediately after the injection, pressure p.sub.2(t.sub.1) at the start of the free injection and pressure p.sub.2(t.sub.2) at the end of the free injection for data-based model 200. Pressure p.sub.2(t.sub.1) and pressure p.sub.2(t.sub.2) are thus determined at points in time, which describe an undisrupted drop (in particular, at the lowest flow resistance) of pressure p.sub.2 in the high pressure accumulator (start, end).

[0052] A data structure of the model data for creating a data-based model, in particular coking model 200, is schematically shown in FIG. 5, which may be used in a method according to the present invention in one preferred specific embodiment. Model 200 obtains as input variables E, for example, instantaneous pressure p as well as instantaneous temperature T of the fuel in the high pressure accumulator and, in addition, the required parameters or variables indicated here with X.sub.1,1, . . . , X.sub.2,1, . . . , X.sub.3,1, . . . and X.sub.4,1, . . . required by the various methods or approaches that have been explained in greater detail with respect to FIGS. 3 and 4. Geometric data or manufacturing data indicated here with Y.sub.1, . . . may also be taken into consideration.

[0053] In model data 200, various data sets, represented by way of example are lines 1 through n, may be taken into consideration or combined with one another in a suitable manner, so that a measure VK results in each case as output variable A for the variable characterizing the flow rate and thus also for the coking. This measure VK may, for example, be indicated in percent of nominal or original flow rate. If needed, the input variables or the corresponding parameters may be viewed as relative to a nominal injector.

[0054] A comparison for the accuracy in a method according to the present invention in one preferred specific embodiment and in a method not according to the present invention is schematically represented in FIG. 6. For this purpose, an estimated or calculated flow rate Q.sub.pred is plotted over a real flow rate Q, the dashed diagonal indicating Q.sub.pred=4, i.e., the ideal case.

[0055] The points represented with a circle correspond in this case to a flow rate estimated or calculated (only) with the aid of one approach while determining the return point in time and closing duration of the nozzle needle, the points represented with a cross or an X correspond, purely by way of example and for illustrating the tendency in a method, with the explained data-based model for determining the flow rate or the coking. In this case, it is clearly apparent that as a result, the accuracy is significantly greater than before.