DE10011690A1 - Fuel injection adaption method for multi-cylinder internal combustion (IC) engine - requires storing first correction factor for basic injection values and then applying these values during stratified-lean operating phases for correcting basic injection values - Google Patents

Abstract

An adaption method for controlling the injection of a multi- cylinder internal-combustion (IC) engine, which is operated phase-wise stoichiometric and lean, involves a first stage (a) of continuously controlling the injection for each cylinder in stoichiometric and/or homogenous-lean operating phases, so that each cylinder on average is operated with stoichiometric or the required homogenous-lean mixture, where a first correction factor is continuously determined and stored, for the basic injection values, and which restores or reproduces the deviation of the actual injection from the desired injection. In a second stage (b), in the stratified-lean operating phases, control of injection for each cylinder is carried out so that each cylinder generates a given torque or that the running smoothness of the engine is maximal. A correction of the basic injection values then follows, during which the first correction factor stored finally in stage (a) is used.

Description

The invention relates to an adaptation method for control the injection of a multi-cylinder internal combustion engine, the phases stoichiometric, Lambda-1 controlled and lean is driven.

To know the fuel consumption of Otto internal combustion engines To reduce ter, internal combustion engines come with lean Combustion is increasingly used. With such a ma General operating mode is between two basic loading Differentiated modes of operation.

In a lower load range, the internal combustion engine is used a heavily layered cylinder load and high air operated excess (hereinafter as stratified-lean Operation designated). This will u. a. through a late one injection into the compression stroke just before the ignition point reached. The internal combustion engine is avoided of throttle losses largely when the throttle valve is open operated.

The internal combustion engine becomes lean in an upper load range and operated with homogeneous cylinder charge (hereinafter referred to as homogeneously lean operation called). The injection takes place already during the intake stroke to ensure thorough mixing to get fuel and air. The air mass sucked in is according to the requested torque, which at for example, requested by a driver on an accelerator pedal is set via a throttle valve.

Finally, the internal combustion engine can also be used with stoichio metric fuel-air mixture are operated (in fol referred to as stoichiometric operation). Doing so the required amount of fuel from the inducted in a known manner  Combustion air mass taking into account the Speed calculated and if necessary via a lambda Regulation corrected.

The homogeneous, lean operation and the stoichiometric operation are hereinafter referred to as "homogeneous operation" summarized.

Fuel injectors naturally have a certain one Deviation of their actual behavior from the specified target Behavior on. This deviation can be due to manufacturing tolerances things, or result from changes in the company, for example through deposits. It is therefore known in stoichiometric operation a so-called cylinder To carry out equality in the cylinder-specific Un Differences of the injection valves can be compensated adaptively. By correcting the control of the respective Injectors ensure that each cylinder is exactly in the stoichiometric operation with lambda 1 control is running. Each after tolerance or age-related deviation that the because injection valve shows, this equality can be a More or less amount of fuel mean that when loading drove the respective injection valve as a correction must be placed.

This cylinder equation is for direct injection Internal combustion engines are of particular importance since their one spray valves directly into the combustion chamber of the internal combustion engine protrude and are therefore particularly strongly influenced by aging are thrown.

It is an object of the present invention to provide a method admit that with an internal combustion engine, both in stoichiometric as well as lean operation, an A daption of the injection control is achieved to change against the injection valves in both stoichiometric and balance in lean operating phases.  

This object is characterized by that in claim 1 Invention solved.

The invention is based on the knowledge that in history tet-lean operation for the behavior of the internal combustion engine essentially the beam characteristic of that of an on spray valve emitted jet is determining. Are there individual changes in injector characteristics in stratified-lean operation mainly relevant to torque, whereas in homogeneous operation (both homogeneously lean and also stoichiometric) of the internal combustion engine mainly e are relevant to the mission. According to the invention is therefore a be Known λ equality in homogeneous operation of the internal combustion engine machine performed, a first correction factor to change change of predefined basic injection values for each injection valve determined and saved. With this first correction Door factor is reached that the respective injectors all show the same actual behavior; tolerance or old Deviations in the delivered fuel mass due to the operation are balanced.

If the internal combustion engine now changes to the stratified lean business, so there will also be equality carried out, now no longer a stoichiometric or homogeneous-lean mixture for the individual cylinders is rend, but that emitted by the respective cylinder Torque; one therefore speaks of torque Equality. To determine the cylinder-specific cor Correction factors of the torque equalization is from respective last stored first correction factor of previous homogeneous operating phase, d. H. the first te correction factor is now used for the stratified-lean Be used drive, with an additional determination or adap tion of a second correction factor takes place, the specific for the stratified and lean operation and together with that first correction factor is used. Starting from these  Values are then carried out using an independent procedure daption of the second correction factor in lean operation.

Because in homogeneous operation it is primarily the force injected substance mass, in the stratified-lean operation but essentially Chen determine the beam characteristics for the behavior the internal combustion engine, the second correction factor, that in the adaptation of a stratified-lean operating phase was determined, hardly on the λ equality in homo operation. Therefore, the λ equality in homogeneous operation when changing the Operating mode from stratified-lean operation to homogeneous Operation again with that in the adaptation algorithm of the ge stratified lean operation unchanged first correction factor that is the result of the adaptation in the previous homo received operating phase, continued and the last value of the second correction factor for the homogeneous Operating phase not used. So there are two adaptions algorithms independent, one for homogeneous operation and one for the stratified and lean operation.

As a target for equal torque in history Tet-lean operation can preferably ensure the smoothness of the burning serve engine. You can do this using, for example of a knock sensor detect the smooth running cylinder selectively and Injection duration and / or start of injection for the individual Change the injection valves appropriately so that they run smoothly increases. Can one in stratified lean operating phases in ge operating conditions do not grasp the smoothness like it for example with strong dynamics of the internal combustion engine Case, it is possible to adapt the second Cor suspension factor.

Of course, the deviation of the actual behavior of an on spray valve from its target behavior not in every phase be the same as the internal combustion engine. For example, think bar that the deviation depends on the fuel pressure. It is  therefore possible in a further training, the cylinder individual correction factors of the λ and / or the torque Make equality dependent on operating parameters. Instead of a single first and second cor to save the correction factor is then for a given Operating parameter classification according to several first and Store second correction factors, for example in suitable ones maps.

Have the separate first and second correction factors further the advantage that the adaptation algo acting on them rithms in homogeneous and stratified-lean operation long can be interpreted sam. In homogeneous operation only works the first correction factor, and only this is adapted, in stratified and lean operation act first and second correction door factor, but only the second correction factor is determined by A daption changed.

Advantageous developments of the invention are the subject of subclaims.

The invention is described below with reference to the Drawing explained in more detail in an embodiment. In the drawing shows:

Fig. 1 is a schematic representation of an internal combustion engine with direct injection and

Fig. 2 is a flowchart of a method for adapting the control of fuel injectors of the engine in FIG. 1.

Fig. 1 shows a schematic representation of an internal combustion engine with gasoline direct injection, which can be operated with both stoichiometric and lean fuel-air mixture. For reasons of clarity, only those components of the internal combustion engine that are necessary for understanding the invention are shown; in particular, only one cylinder of a multi-cylinder internal combustion engine is shown.

The internal combustion engine has a piston 10 which delimits a combustion chamber 12 in a cylinder 11 . In the combustion chamber 12 , an intake duct 13 opens at an inlet valve 14 through which the combustion air flows into the combustion chamber 12 . An exhaust valve 15 connects the combustion chamber 12 to an exhaust tract 16 , in the wide ren course of which there is an oxygen sensor in the form of a broadband lambda probe 17 and a NOx storage catalytic converter 18 with a three-way pre-catalytic converter (not shown).

Using the signal from the lambda probe 17 , the fuel-air mixture is regulated / controlled by a control unit 21 in accordance with the target specifications in various operating modes of the internal combustion engine. For example, a known lambda control takes place in stoichiometric operation.

For such a lambda control, there is a further lambda probe 32 downstream of the NOx storage catalytic converter 18 , which is used for control and setpoint control. In this case, the oxygen probe is a binary lambda probe 32 (two-point lambda probe) which exhibits a jump characteristic with a lambda value of λ = 1. A NOx sensor can also be used instead of the lambda probe 32 . Furthermore, there is usually a temperature sensor 33 in the exhaust tract.

The NOx storage catalytic converter 18 is used in order to be able to comply with the exhaust gas limit values required with respect to NOx compounds when the internal combustion engine is operating leanly. Due to its coating, it adsorbs the NOx compounds in the exhaust gas that it generates during lean combustion.

Exhaust gas recirculation is provided to reduce the NOx emissions that occur especially in internal combustion engines with direct injection in stratified lean operation. By adding exhaust gas to the fresh air drawn in, the temperature of the combustion is reduced, which at the same time reduces the NOx emissions. Therefore, an exhaust gas recirculation line 19 is led from the exhaust tract 16 upstream of the NOx storage catalytic converter 18 to the intake duct 13 , which opens between a throttle valve 20 and the inlet valve 14 into the intake duct. In the exhaust gas recirculation line 19 , a controllable valve 22 is connected, which is usually referred to as a gas recirculation valve. By controlling the Ven valve 22 , the amount of recirculated exhaust gas can be adjusted.

The combustion air for the cylinder 11 flows into the intake duct 13 via an air mass meter 23 . The throttle valve 20 arranged therein is an electric motor-controlled throttle element (E-gas system), the opening cross section of which can be influenced by the control unit 21 in addition to being actuated by a driver (driver pedal position). This can, for example, reduce disruptive load change reactions. In addition, the throttle valve 20 is almost completely opened by the control device 21 in the stratified-lean operation. Furthermore, the control unit 21 ensures, by appropriate intervention on the throttle valve 20, a smooth transition from sto chiometric to homogeneously lean and from there to the lean operation.

Finally, there is a temperature sensor 24 in the intake duct 13 , which is connected to the control unit 21 . Of course, the temperature sensor 24 can also be integrated in the Luftmas senmesser 23 .

In the combustion chamber 12 protrude a spark plug 25 and an injection valve 26 , which is fed for injection with fuel from egg nem high pressure accumulator 27, which is part of a known fuel supply for gasoline direct injection. The control unit 21 is finally connected to a knock sensor 28 , which detects mechanical vibrations on the housing of the internal combustion engine and emits a corresponding signal. The speed of the internal combustion engine is sensed via a sensor 29 which scans the crankshaft or a sensor wheel attached thereto. Further control parameters necessary for the operation of the internal combustion engine, for example accelerator pedal position, signals from temperature sensors etc. are also supplied to the control device 21 and are generally identified in FIG. 1 by the reference symbol 30 .

In the control unit 21 , a block 31 for torque determination and monitoring is finally provided, the function of which will be explained later.

Furthermore, the control unit 21 is connected to a memory 34 , in which various threshold values TQI_SW1, TQI_SW2 as well as at least the maps KF1 and KF2 are stored, the meaning of which will be discussed below.

The control unit 21 determines, depending on the operation, whether the internal combustion engine is to be operated stoichiometrically, homogeneously lean or historically lean.

In each operating mode, the control unit 21 constantly determines the control data for the injection valve 26 , that is to say the start of injection and the duration of injection or the end of injection. The start of injection is related to the crankshaft position, which the control unit 21 is known by means of the sensor 29 . In order to compensate for age and production tolerance-related individual deviations of the individual injection valves 26 in a multi-cylinder internal combustion engine, the control unit 21 carries out an adaptation process, the flowchart of which is shown in FIG. 2, in which the reference symbols beginning with S denote steps of the process sequence.

Corresponding variables are initiated in a step S1 alized. In particular, the map KF1 is either with Default values, or with the last execution values determined during the adaptation process.

In a step S2, it is then queried whether the internal combustion engine is in the homogeneous operating mode (λ = 1). If this is the case, the route continues with the "+" sign. If the internal combustion engine is not in the homogeneous operating mode, the branch marked with a "-" sign is continued. This query is necessary if the adaptation process runs as an independent process in control unit 21 . If, on the other hand, it is integrated in the operating mode control, the query in step S2 can be omitted, since it is then always known which operating mode is present.

In the case of homogeneous operation, the signal of the lambda probe 32 is recorded individually for each cylinder in a step S4. This cylinder-specific recording makes it possible to assess which mixture each cylinder receives on average. The internal combustion engine is operated with the currently applicable control values for the injection. The currently valid control values are composed of a basic control value and a current value of a first correction factor to be described from the map KF1. Subsequently, a query is made in step S5 as to whether the operating mode has been changed in the meantime. If this is the case, the system jumps back before step S2, otherwise the process continues in the branch labeled "-".

Then, in step S6, it is next checked whether it turns out recognize the cylinder-specific detection in step S4 leaves that all cylinders with the target mixture, at stoichio metric operation were operated on average with λ = 1.  

If this is the case, is looped before step S4 jumped back.

If the query in step S6 shows that individual cylinders have not been supplied on average with the target mixture by their injection valves 26 , a fuel quantity correction is calculated cylinder-selectively in step S7. The amount of fuel to be metered to the cylinders via their injection valves 26 is corrected to the target mixture. For cylinders that have been operated with a mixture that is too rich, a reduced fuel quantity is therefore calculated; for cylinders that were operated with a mixture that was too lean, a fuel increase.

This fuel quantity correction is the first mentioned above Correction factor. It is entered in map S8 in step S8 sets.

Then jump back before step S4. In step S4, the control unit 21 is then instructed to take into account the corresponding fuel quantity corrections of the map KF1 when driving the injection valves 26 . This will usually be done by reducing or lengthening the injection period accordingly. Due to the sequence of these steps, cylinder equality is achieved. As mentioned, the loop is only jumped out in step S5 if there is an operating mode change.

If the internal combustion engine runs in stratified, lean operation, the equalization by adapting the injection valves 26 cannot take place with steps S4 to S8, since then the injected fuel mass is no longer predominantly determining the behavior of the internal combustion engine, but also the jet characteristics are essential is to be taken into account. Therefore, the first correction factor, that is to say the excess and reduced fuel quantity of the map KF1, can no longer be used alone. Rather, an independent, additional adaptation for torque equalization in ge stratified lean operation of the internal combustion engine is necessary. For this reason, in lean operation of the internal combustion engine, a further map KF2 with a second correction factor is first accessed in step S9. To equalize the torque, the injection is carried out with two correction values, the first correction value, which remains unchanged during the historically lean operating mode, and the second correction factor, which is changed by adaptation.

Then the injection with the currently valid On tax values made. These consist of a trigger base value, the first correction factor and the current one Value of the second correction factor from the map KF2 together men.

Then, in step S10, the smooth running is cylinder-selective. This takes place in the above-mentioned block 31 of the control device 21 by suitable evaluation of the signal of the knock sensor 28 in order to detect the torque output by each cylinder. This block 31 can for example also fall back on the signals of a torque sensor (not shown in FIG. 1).

The detection in step S10 provides the difference of the torques delivered to the individual cylinders.

Subsequently, in step S11 it is again queried whether a Operating mode change is present. If this is the case, will Step S2 jumps back, otherwise step S12 continued.

This step S12 checks whether the difference of the Torques given torques below a threshold lies. Depending on the operating mode, this can be the Threshold value TQI_SW1 for the case of homogeneously lean operation bes or the threshold TQI_SW2 stratified for the case-  act lean. If the difference is less than Threshold for all cylinders is reset before step S10 jumped, otherwise proceed to step S13.

In step S13, the second correction factor for the consideration of the jet characteristic of the injection valve 26 is updated cylinder-selectively. This adaptation of the two-th correction factor is based on a torque equalization of the cylinders 11 . The second correction factor adapted or changed in this way is entered in the map KF2 for each cylinder.

Now the injection takes place with corrected values. The injection correction can be an injection duration change, but an injection start correction or a combination of the two is also possible. Both correction factors are used for correction. In step S14, the control unit 21 is instructed to take the second correction factor of the map KF2 together with the unchanged first correction factor from the map KF1 into account when actuating the injection valves 26 . Then jump back before step S10.

The adaptation of the control of the injection valves 26 thus uses the first correction factor from the λ equality in stratified-lean operation of the internal combustion engine, but not the second correction factor in homogeneous operation. This is due to the fact that the results of the λ equality for homogeneous operation can be applied to the torque equation for stratified lean operation, because differences in injected fuel mass are taken into account in λ equality in homogeneous operation be valid both there and in ge-lean operation of the internal combustion engine. The second correction factor, which is adapted in the torque equalization in the stratified-lean operation, compensates for a change in the jet characteristic of the injection valves 26 , for example due to coking. However, these differences in the jet characteristics of the injection valves 26 are not or only barely relevant in the homogeneous operation of the internal combustion engine, which is why the results of the torque equalization in the adaptation process in the stratified lean operation of the internal combustion engine do not provide the first correction factor for the λ equalization in the adaptation process homogeneous operation of the internal combustion engine may react.

Claims (6)

1. Adaptation method for controlling the injection of a multi-cylinder internal combustion engine which is operated stoichiometrically and lean in phases, in which method the following stages are carried out:

• a) in stoichiometric and / or homogeneously lean operating phases, the control of the injection is effected continuously for each cylinder in such a way that each cylinder is operated on average with stoichiometric or desired homogeneous-lean mixture, a basic correction factor being continuously determined for basic injection values and is stored, which reflects the deviation of an actual injection from the target injection, and
• b) in stratified, lean operating phases, the control of the injection is effected continuously for each cylinder in such a way that each cylinder generates a predetermined torque or that the smooth running of the internal combustion engine becomes maximum, a correction of basic injection values being carried out, the last being in stage a) stored first correction factor is used.

2. The method according to claim 1, characterized records that in stage b) a second correction factor is won, which together with the last in step a) ge stored first correction factor is used and the Deviation of the actual injection from the target injection for reflects the stratified lean operation. 3. The method according to any one of the preceding claims characterized by that after an Ü transition from a stratified, lean operating phase to one stoichiometric or homogeneously lean operating phase at the Control of the injection in stage a) with the last ge stored value of the first correction factor continued becomes.   4. The method according to any one of the preceding claims characterized in that in stage a) the first correction factor the deviation of the injected Fuel mass reflects. 5. The method according to claim 2, characterized records that in stage b) the second correction factor is adapted, but the first correction factor remains unchanged remains. 6. The method according to any one of the preceding claims characterized by that the first and / or second correction factor depending on operating parameters is selected and in an operating parameter-dependent map is filed.