GB2326742A

GB2326742A - Determining, predicting and regulating the fuel pressure in an internal combustion engine

Info

Publication number: GB2326742A
Application number: GB9813564A
Authority: GB
Inventors: Hartmut Bauer
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 1997-06-24
Filing date: 1998-06-23
Publication date: 1998-12-30
Anticipated expiration: 2018-06-23
Also published as: GB9813564D0; DE19726757A1; FR2764943B1; JPH1162671A; DE19726757B4; FR2764943A1; GB2326742B; JP4250222B2

Description

1 2326742 CONTROL OF OPERATION OF AN INTERNAL COMBUSTION ENGINE The

present invention relates to a method of controlling operation of an internal combustion engine, especially of a motor vehicle.

In a known method of controlling operation of a directly injected diesel or petrol internal combustion engine, each combustion chamber of the engine is associated with a respective injection valve by which fuel is injected directly under pressure into the respective chamber and a pump, by which the fuel is pumped to the valves, is provided for the production of the pressure acting on the fuel. The pressure produced by the pump is usually not constant. If, for example, the pump is a piston pump, then the pressure fluctuates in dependence on the number and setting of the pistons. This varying pressure, which is produced by the pump and acts on the fuel, is measured by a pressure sensor.

The pressure acting on the fuel during injection is significant for inter alia dimensioning of the mass of fuel to be injected into the combustion chamber. Thus, for example, a shorter injection duration is required for the same mass of fuel to be injected at a higher pressure, whereas the injection valve must be kept open longer in the case of a lower pressure.

The problem now exists in that on the one hand the pressure acting on the fuel during the respective injection phase or cycle is not known in advance and on the other hand this pressure must be taken into consideration for the dimensioning of the mass of fuel to be injected. In the case of known methods, this problem is solved by the foreseeable pressure during an injection cycle being deduced from pressure values measured in the past and the mass of fuel to be injected then being corrected in dependence thereon. However, this results in rather inaccurate dimensioning of the mass of fuel to be injected.

There is thus a need for a method by which a more precise ascertaining of the pressure expected to act on the fuel is possible for an injection cycle.

According to a first aspect of the present invention there is provided a method of operating an internal combustion engine, especially of a motor vehicle, having several combustion chambers, in which fuel is injected under a pressure directly into the combustion chambers of the engine and in which the pressure acting on the fuel is measured, characterised in 2 that the pressure foreseeably acting on the fuel during the next injection of fuel into one of the combustion chambers is ascertained from a mean value and a correction value.

According to a second aspect of the invention there is provided an internal combustion engine, especially for a motor vehicle provided with several combustion chambers, each of which is associated with a respective injection valve by which fuel can be injected directly into the associated combustion chamber, with a pump for the production of a pressure on the fuel fed to the injection values, with a pressure sensor for measuring the pressure acting on the fuel and with a control device for the control and/or regulation of magnitudes influencing a combustion of the fuel in the combustion chambers, characterised in that the control device ascertains the pressure foreseeably acting on the fuel during the next injection of fuel into one of the combustion chambers from a mean value and a correction value.

Due to the splitting-up of the ascertaining of the foreseeable pressure into a mean value and a correction value, the mean value can also be used separately. Thus, for example, the mean value can be used for the control and/or regulation of the pressure acting on the fuel.

Preferably, a regulating loop is formed to regulate the mean value to a desired target value. This has the consequence that the mean value changes only relatively slowly. If the mean value is ascertained from preceding measurements of the pressure, it can be presupposed, in view of the slow change, that the mean value ascertained in this manner substantially agrees with the actual mean value during the next injection.

Through the use of the mean value for the control andlor regulation of the pressure acting on the fuel, it is thus achieved that the actual, not yet measurable mean value of the next injection cycle corresponds substantially with the mean value ascertained by the preceding measurements. By virtue of this predictability, the mean value is particularly suitable for utilisation in the ascertaining of the mass of fuel to be injected into a combustion chamber.

When the mean value is drawn on for ascertaining the mass of fuel to be injected into a combustion chamber, then the total pressure acting on the fuel is not yet taken into consideration. The deviations from the mentioned mean Value must still be taken into account. These deviations include pressure fluctuations, for example in dependence on 3 the settings of the pistons in the case of a piston pump. These deviations, however, correspond to the correction value. The correction value in that case represents the deviation, which changes rapidly from injection to injection, of the pressure, which foreseeably acts on the fuel, from the mean value.

This has the consequence that the mass of fuel to be injected can be ascertained in dependence on the mean value and the correction value. The mean value changes only slowly so that to that extent substantially no errors can arise during ascertainment of the mass of fuel to be injected. The correction value, thereagainst, changes rapidly with each injection, but these changes are small by comparison with the mean value.

The correction value can be ascertained from measurements of the pressure acting on the fuel. Since only the pressure for preceding injections can be measured and the foreseeable pressure during the next injection must be deduced therefrom, it is possible that an error occurs in the ascertainment of the correction value. In that connection, because the correction value is subject to more rapid change it is more likely to suffer from an error. Thus, when an error arises in the ascertainment of the mass of fuel to be injected, this error is usually based substantially on an error during ascertainment of the correction value, rather than of the mean value. However, since the changes in the correction value, as explained, are small by comparison with the mean value, the error tends to be small.

By contrast to the state of the art, in which the entire measured pressure value can be the basis for an error in the ascertainment of the mass of fuel to be injected, an error in the case of a method exemplifying the invention is, by reason of the splitting-up into a mean value and a correction value, based only on the changing small correction value, not on the mean value. Thereby, it is achieved that a possible error during ascertainment of the mass of fuel to be injected is generally smaller than for the state of the art. This results in a more exact dimensioning of the mass of fuel to be injected and thereby a better combustion in the combustion chambers. Resulting therefrom are then advantages such as a smoother running of the engine, a lower fuel consumption and/or a reduction in exhaust emission.

Further advantages are that the influence of the different conveying performances of the individual cylinders of the pump conveying the fuel can be taken into consideration and 4 corrected. Similarly, inaccuracies in angle of the installation of the pump with respect to the camshaft can be taken into consideration and corrected, so that an angularly exact installation of the pump is not critical.

In an advantageous example, the correction value is ascertained either from at least one measurement of the pressure acting on the fuel during the preceding injection of the fuel into the same combustion chamber or from at least one measurement of the pressure acting on the fuel during the preceding injection of fuel into the preceding combustion chamber. The correction value is thus either dependent on the most recent injection into the same combustion chamber or on the most recent injection into any one of the combustion chambers In both cases, the pressure acting on the fuel during each injection is measured and stored. For the computation of the mass of fuel to be injected into the next combustion chamber, either the stored value of the same combustion chamber or the stored value of the combustion chamber having the last injection is used as the correction value and thereby as pressure foreseeably acting on the fuel.

In a further advantageous refinement, the correction value is ascertained from the measurement of the pressure acting on the fuel in about the middle of the injection or from the measurement of the pressure acting on the fuel, in particular, shortly before or after the injection with the aid of an averaging process or the like or from the measurement of the pressure acting on the fuel in particular shortly before the injection with the aid of an adaptation process or the like. Theoretically, the continuous measurement of the pressure acting on the fuel during the entire injection, thus during the entire injection duration, would be the most accurate basis for the ascertainment of the correction value. In practice, however, this is not feasible. Appropriate alternatives are thus measuring the pressure before and after the injection and ascertaining the correction value therefrom. These possibilities can in that case be used individually or together. The correction values thereby ascertained on the one hand have a high accuracy and reliability and on the other hand can be ascertained with tolerable expenditure, in particular with acceptable computing effort.

In a further advantageous refinement, the mean value is used for the control and/or regulation of the pressure acting on the fuel. The mean value in that case represents an actual value which can be regulated towards a target value by control and/or regulationThe mean value can thus be used for ascertaining the mass of fuel to be injected. Due to the twofold use of the mean value, on the one hand for control andlor regulation of the pressure acting on the fuel and on the other hand for ascertainment of the mass of fuel to be injected inter alia in dependence on the pressure acting on the fuel, the ascertainment of the fuel mass can be optimised by an appropriate control andior regulation of the pressure.

It is in that case particularly advantageous if the target value of a desired pressure acting on the fuel during the next injection of fuel into one of the combustion chambers is ascertained in dependence on the engine rotational speed and/br load andlor on layer andior homogeneous operation of the engine. In this manner, the ascertainment of the mass of fuel to be injected can be interlinked particularly well with the control andlor regulation of the pressure acting on the fuel. Moreover, it can be achieved thereby that the mean value changes substantially only slowly. This brings the already mentioned advantage that substantially no errors result from the mean value as such during the ascertainment of the mass of fuel to be injected.

In a further advantageous example, the pressure foreseeably acting on the fuel during the next injection of fuel into one of the combustion chambers is used for the control and/or regulation of the mass of fuel to be injected into the combustion chamber. As already explained, the error possibly then arising in the ascertained pressure is smaller than in the case of the state of the art. The control and/or regulation of the mass of fuel to be injected is thus more accurate. This has the consequence that the engine may have a quieter and smoother running, thus be less susceptible to abrupt transitions, and may use less fuel and emit fewer noxious gas constituents. The provision of an electrical storage medium, which is usable in a control device of an internal combustion engine, in particular of a motor vehicle, can be of special significance. In that case, a program, which is capable of running sequentially on a computing device, in particular on a microprocessor, and suitable for the performance of the method according to the invention, can be stored on the electrical storage medium.

An example of the method and an embodiment of the engine of the present invention will now be more particularly described With reference to the accompanying drawings, in which:

6 Fig. 1 is a schematic block diagram of control means for controlling operation of an internal combustion engine; and Fig. 2 is a diagram showing signals generated in the control means.

Fig. 1 shows control means for a four-cylinder internal combustion engine with four combustion chambers. Each combustion chamber is associated with an injection valve, by which fuel can be injected directly into the chamber. Preferably, petrol is provided as fuel, but diesel can also be used. For injection of the fuel into the chambers, a pressure is exerted on the fuel by a pump which has, for example, three cylinders. For measurement of the pressure acting on the fuel, a pressure sensor is arranged for example in the region of the pump. For influencing the pressure acting on the fuel, a pressure control valve is provided, in the open setting of which the pressure, for example, fails.

The mass of fuel injected by an injection valve results from inter alia the duration of an injection cycle, during which the injection valve is open, as well as from the pressure acting on the fuel during the cycle. If a certain mass of fuel is to be injected into a combustion chamber, the required injection duration must be ascertained in dependence on the pressure acting on the fuel during this injection. If the pressure is greater, the injection duration becomes smaller, and conversely.

The ascertainment of the duration of a particular injection must be computed in advance. For this purpose, the pressure acting on the fuel during the injection must also be taken into consideration in advance. This future pressure is not, however, known in advance and therefore has to be computed in another manner.

In Fig. 1, the pressure measured by the pressure sensor and acting on the fuel is denoted by "up" and the number of the respectively relevant combustion chamber is denoted by "Cyl". The pressure "up" and the combustion chamber number "Cyl" are processed by a block 1 for ascertaining the pressure acting on the fuel. Measured pressure "up" is linearised andlor standardised by a function 2 in the block 1. By the function 2, a pressure p is produced at the output, which is fed to a switch 3, a mean value formation stage 4 and a correction value formation stage 5. The stage 4 produces a mean value M, which is fed to the switch 3. The stage 5 produces correction values Delta CylA, Delta Cy12, Delta 7 Cyl.3 and Delta Cyl.4, which are fed to a control and/or regulation block 6, by which the mass of fuel to be injected into a combustion chamber can be influenced.

The mean value M is formed to be dependent on time and/or angle. In particular, the mean value M is computed on the basis of a detection, which is equidistant in time and/or angle, of the pressure p.

In Fig. 2 it is presupposed that the three-cylinder pump and the fourcylinder engine are matched to each other in such a manner that running of the pump through an angle of 36011 corresponds exactly to running of the engine through a camshaft angle of 3600. This has the consequence that the pressure produced by the pump has three waves over 3600, whilst the engine has four injections during this operation.

If other numbers of cylinders are provided in the engine and/or the pump or if a step-up or step-down gear is present between the engine and the pump, this can be taken into consideration with the aid of a correction function or a correction characteristic values field.

In the illustrated example, it is possible according to Fig. 2 for the mean value M to be computed from four measurement values M1, M2, M3 and M4, which correspond with the pressure p in four successive instants in each case shortly before an injection. The injection is indicated in Fig. 2 by the injection durations tiCyl.1 to tiCyl.4 for the four cylinders. The four successive instants each have a spacing of 900 of camshaft angle and are thereby equidistant in angle.

The four measurement values M1, M2, M3 and M4 are added up and averaged by the mean value formation stage 4, thus divided by four, whereby the mean value M as illustrated in Fig. 2 results.

In corresponding manner, it is also possible to detect the pressure p at constant time intervals and process it into a mean value M. Moreover, it is possible to combine angledependent and time-dependent ascertainment of the mean value M. This can be effected, for example, by switching to and fro from a time-dependent ascertainment to an angledependent ascertainment in dependence on the engine speed.

8 In case of normal changes in the pressure p, thus when dpIdt is smaller than a preset threshold value S, the switch 2 is disposed in the illustrated setting so that the mean value M is passed on as actual value pi of the pressure acting on the fuel. In the case of strong fluctuations of the pressure p, thereagainst, for example after starting, thus when dpIdt is greater than the threshold value S, the switch 2 is switched over by a block 7 and the pressure p is passed on directly as actual value pi of the pressure acting on the fuel.

The actual value pi of the pressure acting on the fuel is fed to a regulation stage 8, by which this value is regulated towards a target value ps. The target value ps is in that case preset in dependence on whether the engine is in homogeneous operation or a layer operation or in that case during the start or after the start. This is achieved with the aid of the selecting bits B_LAYER and START, which switch over respectively associated switches 9 and 10.

Homogeneous operation denotes that operational state of the engine in which the fuel is injected into the combustion chamber during the induction phase. In layer operation, the fuel is injected into the combustion chamber during the compression phase.

If the engine is in homogeneous operation after the start, then a value phomogenoustarget is preset as the target value ps. If the internal combustion engine is in layer operation after a start, then a value player-target is preset as the target value ps. If the engine is in homogeneous operation during the start, then a value p-starthomogenoustarget is preset as target value ps, and if the engine is in layer operation during the start, then a value p-start-layer-target is preset as the target value ps.

The values p-homogenous-target and p-layer-target are ascertained with the aid of characteristic value fields 11 and 12 from the rotational speed n of the engine and the load m applied to the engine.

After a comparison point 13 between target value ps and actual value pi, the difference between the target and actual values is fed to a proportional-integral regulator 14, the regulating constants kp and ki of which are formed by the blocks 15 and 16 by appropriate functions or characteristic value fields. The regulator 14 also has a minimum value MIN and a maximum value MAX and is resettable by means of a signal Reset, for example in the case of standstill of the motor or in the case of a large change in the target value. The

9 output value of the regulator 14 is added at the point 17 to the target value ps in order thereafter to be adapted with the aid of a function or a characteristic values field 18 at the pressure control valve influencing the pressure. With the aid of a further characteristic values field 19, the regulator value is subjected, at a multiplication point 20, to a voltage correction in dependence on the battery voltage UBatt of the motor vehicle. Finally, the signal produced in this manner is reshaped with the aid of the block 21 in dependence on the frequency fDSV into a keying ratio OUT, by which the pressure control valve is then controlled in drive.

In case of normal changes in the pressure p, thus when the mean value M corresponds vAth the actual value pi, the mean value M is thus regulated by the regulation stage 8 to the target value ps. Since the fluctuations of the pressure p, as presupposed, are smaller than the preset threshold value S, the mean value M changes only slowly. The change is in that case dependent by way of the regulation 8 on the change in the target value ps and thereby dependent on the speed and load.

The slow change in the mean value M is also evident from Fig. 2, in which the mean value M is illustrated as about constant over a passage of the camshaft through 3600.

The cylinder-specific correction values Delta Cyl. 'I to Delta CylA can be ascertained by the correction value formation stage 5 in different ways. This is explained by way of example in the following by reference to the correction value Delta CylA, but the same applies to the other correction values.

It is possible that the correction value Delta CyLl is ascertained by the pressure p being detected and added up or integrated by the stage 5 over the entire injection duration tiCyLl of the cylinder CyLl. This represents the most accurate ascertainment of the correction value.

It is also possible for the correction value to be ascertained by the pressure p being detected in about the time centre of the injection duration and in a given case weighted by a factor. This results in a correction value, which is relatively accurate when the pressure p remains substantially constant over the course of the injection duration tiCyi. l. Equally, it is possible for the correction value Delta Cyl. l to be ascertained by the pressure p being detected shortly before the beginning of the injection and shortly after the end of the injection. These measurement values are then averaged and, optionally also, weighted by a factor. This correction value is, by reason of the averaging, still relatively accurate when the pressure fails over the course of the injection duration. Moreover, the correction value can be ascertained by the pressure p being detected shortly before the beginning of the injection. The course of the pressure during the injection duration is then deduced from this measurement value with the aid of appropriate adaptation methods.

The above-described possibilities can be applied alternatively or cumulatively.

In all of the described possibilities, the pressure p is detected before and/or during an injection. This injection can be either the last injection into the same combustion chamber or the last injection into any other combustion chamber.

In Fig. 2, the second explained possibility for the same combustion chamber is illustrated by way of example. Thus, for example, the correction value Delta y14 for the next injection of fuel into the fourth cylinder CylA is computed from the pressure p in about the time centre of the injection duration tiCylA of the last injection of fuel into the same fourth cylinder.

The correction values are fed to the control and/or regulation block 6, by which the mass of fuel to be injected into a combustion chamber is ascertained. This mass of fuel to be injected must be ascertained in advance in dependence on a number of parameters of the engine operation, including the pressure foreseeably acting on the fuel during this injection. This pressure foreseeably acting on the fuel during the next injection is ascertained from the mean value M and the respective one of the cylinder-dependent correction values.

If, for example, the pressure acting on the fuel during the next injection into the combustion chamber of the first cylinder CyLl is to be ascertained again, the mean value M is computed in the described manner. The correction value Delta CyLl is computed according to one of the described possibilities, wherein the computation can be based either on the preceding injection, thus preceding by 7201>, into the same first cylinder Cyl. l or on the last injection, thus preceding by 1801>, into another cylinder, thus according to Fig. 2 on the injection into the second cylinder Cy12.

The mean value M and the respectively relevant correction value thus represent the pressure foreseeably acting on the fuel during the next injection. This pressure is computed in advance with the aid of the mean value M and the mentioned correction value. This advance computation of the pressure is then used for ascertainment of the injection durations so that the fuel mass, which is optimum for the respective operational state of the engine, can be injected into the combustion chamber.

As has been explained, the mean value M changes substantially only slowly. To that extent, no errors are to be expected in the advance computation on the basis of the mean value M. The correction value, however, changes more rapidly, as is also evident from Fig. 2. The changes are small by comparison with the mean value M. For this reason, errors are to be expected in the advance computation on the basis of the mentioned correction value. These are, however, rather small in view of the small changes in the correction value.

The entire process can be performed by a control device 22, such as a programmable microprocessor, which is provided with storage devices and other required components and built into a motor vehicle. The control device 22 in that case receives the signals required for the performance of the method from inter alia respective sensors, for example the pressure sensor, and according to the described method generates therefrom the required signals for the control of, for example, activators, which activate the injection valves or the pressure control valve.

12

Claims

1. A method of controlling operation of an internal combustion engine, comprising the steps of injecting fuel under pressure directly into each of a plurality of combustion chambers of the engine in successive injection cycles for that chamber, measuring the pressure acting on the fuel, determining the pressure expected to act on the fuel during the respective next injection cycle for a combustion chamber in dependence on a mean value and a correction value, and utilising the determined pressure value in connection with control of the engine or of a vehicle fitted with the engine.

2. A method as claimed in claim 1, wherein the correction value is ascertained from at least one measurement of the pressure acting on the fuel during injection of fuel into the same chamber during the preceding injection cycle.

A method as claimed in claim 1, wherein the correction value is ascertained from at least one measurement of the pressure acting on the fuel during injection of fuel into the preceding chamber during the preceding injection cycle.

4. A method as claimed in claim 2 or claim 3, wherein said at least one measurement is taken at or in the region of the middle of the respective injection cycle,

5. A method as claimed in claim 1, wherein the correction value is ascertained by an averaging process applied to measurements of the pressure obtained immediately before and after an injection cycle.

6. A method as claimed in claim 1, wherein the correction value is ascertained by an adaptation process applied to measurements of the pressure obtained immediately before an injection cycle.

7. A method as claimed in any one of the preceding claims, wherein the mean value is ascertained from measurements of the pressure acting on the fuel, the measurements being dependent on at least one of time and engine crankshaft rotational angle.

13

8. A method as claimed in claim 7, wherein the measurements dependent on at least one of time and rotational angle are, respectively, equidistant in time and equidistant in angle spacing.

9. A method as claimed in any one of the preceding claims, comprising the further step of utilising the mean value for at least one of control and regulation of the pressure acting on the fuel in said next injection cycle.

10. A method as claimed in any one of the preceding claims, wherein the step of determining is carried out in dependence on a target value for the pressure acting on the fuel in said next injection cycle.

11. A method as claimed in claim 10, wherein the target value is dependent on at least one of the speed and the load of the engine.

12. A method as claimed in claim 10 or claim 11, wherein the target value is dependent on a value indicative of layer operation or homogeneous operation of the engine.

13. A method as claimed in any one of claims 10 to 12, wherein the step of determining comprises comparing the target value with the mean value.

14. A method as claimed in any one of the preceding claims, wherein the step of utilising comprises carrying out at least one of control and regulation of the mass of fuel for said next injection cycle in dependence on the determined pressure value.

15. A method as claimed in claim 1 and substantially as hereinbefore described with reference to the accompanying drawings.

16. An electrical storage medium for control means for an internal combustion engine, the storage medium storing a program for causing computing mans of the control means to perform the method claimed in any one of the preceding claims.

17. A storage medium as claimed in claim 16, the medium having a readonly memory capability.

14

18. A multi-cylinder internal combustion engine provided with a fuel injection installation, which comprises a respective fuel injection valve for direct injection of fuel into each combustion chamber of the engine and pressure-generating means for pressurising fuel to be injected by the valves, and with control means for measuring the pressure acting on the fuel, determining the pressure expected to act on the fuel during the respective next injection cycle for a combustion chamber in dependence on a mean value and a correction value, and utilising the determined pressure value in connection with control of the engine or of a vehicle fitted with the engine.

19. An engine as claimed in claim 18, the control means being operable to ascertain the correction value from a measurement of the pressure acting on the fuel during injection of fuel into the same or the preceding chamber during the preceding injection cycle.

20. An engine as claimed in claim 19, wherein said measurement is taken at or in the region of the middle of the respective injection cycle.

21. An engine as claimed in claim 18, wherein the correction value is ascertained in dependence on measurements of the pressure obtained immediately before and after an injection cycle.

22. An engine as claimed in any one of claims 18 to 21, the control means being operable to utilise the determined pressure valve for at least one of control and regulation of the mass of fuel for said next injection cycle.

23. An internal combustion engine substantially as hereinbefore described with reference to the accompanying drawings.