EP0757168B1 - Method and apparatus for controlling an internal combustion engine - Google Patents

Description

State of the art

The invention relates to a method and a device for controlling an internal combustion engine according to the generic terms of independent claims.

Such a method and such a device for control an internal combustion engine is, for example, from the DE-OS 41 05 740 (US 315,976) known. There is a procedure for additive and multiplicative correction of a map described. A first deviation signal represents an additive and a second deviation signal one multiplicative error. These additive and multiplicative Errors are caused by additive and multiplicative correction factors taken into account in the entire map area.

Such a correction of global correction Maps only provide sufficiently precise correction values if the map is divided into areas in which dominate the additive errors, and in such areas, in which multiplicative errors dominate.

Furthermore, there are also local corrections to a map known in which a correction value for each map base is learned. These corrections are more precise, point out but the disadvantage that discontinuities occur because certain points in a driving cycle are already learned could, while in neighboring points for this yet Opportunity was. For example, it takes a long time to the changed fuel properties after a refueling process are taken into account in the entire map. Because certain Operating points that are only rarely approached occur with these a functional impairment when starting off for the first time on. For example a weak torque or an inadmissible soot emission. This continues until the adaptation has also taken place at this operating point.

Object of the invention

The invention has for its object in a method and a device for controlling an internal combustion engine of the type mentioned at the beginning, as simple as possible and to achieve exact adaptation of a map.

This task is accomplished by the in the independent claims marked features solved.

Advantages of the invention

The procedure according to the invention has the advantage that a simple adaptation of the map is possible there is very precise where it is needed for functionality.

Advantageous and expedient refinements and developments the invention are characterized in the subclaims.

drawing

The invention is described below with reference to the drawing illustrated embodiments explained. 1 shows 2 shows a block diagram of the device according to the invention, FIG. 2 the correction map and Figure 3 different work areas.

Description of the embodiments

The procedure according to the invention is described below on Example of a pump map of a diesel engine described. However, the invention is not based on this Application limited. You can by making minor changes also on other internal combustion engines or other maps be applied. It is also possible that instead of Pump map for other maps, for example one Exhaust gas recirculation map, a corresponding correction is made.

100 denotes a first actuator, which depends on the control signal US supplied to him, not shown Internal combustion engine a certain amount of fuel ascribes. The quantity-determining actuator 100 is operated by a so-called Pump map 105 is supplied with the signal US. The pump characteristic map 105 receives the output signal as an input variable of node 110 forwarded. At first The entry of node 110 stands with a positive sign an output signal MKS of a minimum selection 120.

The minimum selection 120 becomes the output signal MKW a quantity specification 142, which for example a signal FP an accelerator pedal position sensor 144 evaluates, supplied. As The second selection becomes the minimum selection 120 the output signal a smoke control 122 and the output signal of a Torque limit 124 supplied. The smoke control 122 evaluates, for example, the output signal λ of a lambda sensor 130, which detects the oxygen concentration of the exhaust gas, and / or the output signal ML of an air quantity sensor 133 out. The torque limit 124 is particularly one Speed signal N 136 supplied.

The output signal of the minimum selection 120 can be still further Control loops are fed. For example, it becomes one Injection start regulator 140, which depends on this Quantity signal MKS sets the desired start of injection. Furthermore, it can be an exhaust gas recirculation controller 145 or an air flow regulator. This regulator also includes a map in which depending on operating parameters a control signal for controlling a second Stellers 148 is stored. This second actuator influences for example the amount of air sucked in via an exhaust gas recirculation flap. The exhaust gas recirculation controller 145 processes the output signal of the speed sensor 136 and the lambda sensor 130 and / or an air mass meter. Such an arrangement is essentially known.

The second input of node 110 is negative Sign the output signal K of an adaptation 115 fed. The adaptation processes the output signal of a Junction point 155 and the speed signal N des Speed sensor 136 and a fuel quantity signal MK, the is provided by a block 157. The addition point 155 becomes the output signal with a negative sign MCS of the minimum selection and with a positive sign a signal MKI fed to a quantity calculation 150. The Quantity calculation 150 are the lambda signal as input variables λ of the lambda sensor 130 and an air quantity signal MLV an air quantity specification 152 or the air mass sensor 133 fed.

The output signal K of the adaptation 115 is from a correction map 180 provided. The correction map 180 the output signal of a first controller 170, one second controller 172 and a third controller 174 fed. The first controller 170 is above a first switching means 160, the second controller 172 via a second switching means 162 and the third controller 174 via a third switching means 164 with connection point 155 in connection. The switching means 160, 162 and 164 are controlled by an adaptation controller 166 controlled depending on operating parameters. As operating parameters for example, the speed signal N and a quantity signal MK is used.

This facility now works as follows:
Depending on a driver request signal FP, which is detected, for example, with an accelerator pedal sensor 144, block 142 specifies a desired quantity MKW that corresponds to the driver request. This desired quantity is limited to the maximum permissible values depending on the output signal of the smoke limitation 122 and the torque limitation 124. The smoke limitation 122 depends, for example, on the air quantity ML supplied to the internal combustion engine and the lambda value λ, that is to say the oxygen concentration of the exhaust gas.

The torque limit 124 is essentially dependent on the speed. At the exit of the minimum selection is 120 the setpoint MKS for the fuel quantity MKS to be injected on. This variable can be fed to various controllers depending on this size, for example the start of spraying or set the exhaust gas recirculation rate. Furthermore, this variable MKS fed to the so-called pump map 105. The pump map sets the MKS signal with respect to amount of fuel to be injected into a control signal US for the actuator 100 µm, the amount of fuel to be injected sets.

Due to errors and inaccuracies in the area of Fuel metering, in particular of the actuator, can be the case occur that the amount of fuel actually injected deviates from the desired amount of MKS. Is the actual quantity less than the setpoint MKS, so delivers the internal combustion engine does not have the desired torque. is the amount too large, there may be impermissible exhaust gas emissions on.

To avoid these effects, a quantity calculation is used 150 the amount of fuel actually injected MKI determines and in comparison point 155 with the target quantity FMD compared. Depending on this comparison result then a correction quantity K is determined, which is also a correction quantity K is called. With this correction variable in Addition point 110 corrected the target value for the quantity MKS.

In the exemplary embodiment described, the intake air quantity MLV and the lambda value λI of the exhaust gas are used to calculate the quantity 150. The actual amount of fuel MKI injected is calculated using the following formula: MKI = MLV 14.5 * λ I

The air volume signal MLV can be an air volume that corresponds to a Sensor is detected immediately, are used, or that Air volume signal MLV can be based on various operating parameters such as temperature and Pressure of the intake air volume can be calculated.

Instead of calculating the actual MKI volume based on the air volume MLV and the lambda signal can also have other quantities or quantity replacement signals are used. For example can the needle stroke, or the spray duration, the pressure in the Fuel line, torque, exhaust gas temperature or the output signal of a NOX or HC sensor is used become. The lambda value λI of the exhaust gas is usually measured directly with a lambda probe.

By comparing the target quantity MKS in connection point 155 with the actual quantity MKI there is a deviation signal DMK. Alternatively, the output signal of a Lambda controller are used.

The adaptation controller 166 ensures that the controllers 170, 172 and 174 only get a signal when certain Operating parameters are available. As an operating parameter the speed N and the fuel quantity MK are taken into account. The fuel quantity MK is an in the fuel quantity value present to the control device, such as for example the target fuel quantity MKS. Alternatively, you can other quantity signals, such as the quantity MKI, can be used.

If certain operating parameters are available, is optional one of the switches 160, 162 and 164 closed and the deviation signal DMK the corresponding controller 170, 172 or 174 fed. These controllers are preferably integral controllers realized. This is a slow one Control loop, which is the determined difference in quantity between the target quantity MKS and the calculated actual quantity MKI regulates to 0 at a certain operating point.

Because the actual quantity MKI, that of the injected fuel quantity corresponds, and the target fuel quantity MKS same Anyone working with the quantity value MKS can accept values Functions such as the spray start controller or improves or simplifies the exhaust gas recirculation control become.

According to the invention it is provided that the deviation signal DMK only in the vicinity of three operating points are defined by the speed and fuel quantity MK become. Based on these three deviation values three correction values are determined for these three operating points. These three correction values at three operating points define a so-called correction level. This correction level assigns each operating point by a fuel quantity value MK a speed value N is defined, a correction amount K to.

According to the invention, these three points are chosen so that in every functionally important work area of the internal combustion engine an operating point comes to rest. A first area of work is at low speeds and small amounts of fuel given. Exhaust gas recirculation is in this operating range active. In a second work area at low speeds and large amounts of fuel Smoke control active. In a third work area at high speeds and large amounts of fuel Torque limitation.

In each of these operating areas there is a correction value learned. Based on these three correction values then the correction plane is calculated, over which finally one global correction of the pump map is carried out.

In Figure 2 are the three operating points at which the correction values be identified, marked with crosses. The speed N is plotted on a first axis. about the fuel quantity MK is plotted on a second axis and The correction quantity K is plotted on a third axis.

In the first operating point, which is determined by the speed N1 and If the amount of fuel MK1 is determined, a first one results Correction value K1. In the second operating point, defined by the speed N2 and the fuel quantity MK3 become a second Correction value K2 determined. In the example shown the speed values N1 and N2 are the same.

Correspondingly, at the third operating point, is defined by the speed N3 and the quantity MK3 a third correction value K3 determined.

The speed N1 takes for example a value of 1000 min ^-1 and the speed N2 a value of 4000 min ^-1 .

Through these three operating points, which are also called bases one correction level, and the three Correction values K1, K2 and K3 define a level that is indicated by a dash-dotted line. Anyone Operating point, that means any combination from speed value N and fuel quantity value MK a point on the plane and thus a certain correction quantity K assigned.

The controllers 170, 172 and 174 set the correction values K1, K2 and K3 available. Based on these correction values and the known operating points assigned to these values there is a correction level in the correction map 180 is filed. From this correction map 180 can be a correction amount K for each operating point be read out.

A correction value is learned in each of the work areas. This is preferably the mean over several measurements of the deviation signal DMK. Starting from The three correction values K1, K2 and K3 then become a correction plane over which a global correction of the Pump map is done. The high accuracy of the global Correction is right there for functionality is needed.

With this approach, both multiplicative and also corrected additive errors. Areas for separate capture of multiplicative or additive errors no distinction, which simplifies the procedure in the application. Global errors can correct the errors learned quickly and thus in the entire map area without Discrepancies can be quickly compensated.

The selected valid deviation signals DMK are from the controller responsible for the respective work area 170, 172 and 174 averaged continuously. As long as no usable There is a signal for the adaptation, i.e. the corresponding one Operating point has not yet been reached, that takes Output signal of the corresponding controller has the value 0. All Deviation signals, which at an operating point within of a work area and considered valid are reached via the switching means 160, 162 or 164 the associated controller 170, 172 or 174, which has a large integration time has.

The continuously averaged quantity errors can be as Display the correction map for the pump map. The Support points at which the correction values are calculated are preferably in the middle of the work area or in placed the area of the work area whose values are most common occur. The between these three correction values spanned level approaches globally after a relatively short time Measurement time on a completely measured correction map. Since only three correction values are used to calculate the level are necessary, all intermediate values are sufficient Accuracy available very quickly.

This procedure has the advantages that in every operating point after a short measuring time in its vicinity measured additive correction value is present. multiplicative Error rates are indirectly determined by the three Correction values defined plane equation also taken into account. By calculating the level over the entire operating area detects the correction with sufficient accuracy also operating points that are rarely used. discontinuities like a point-by-point pump map, do not occur. The application effort can be greatly reduced become.

In a particularly advantageous embodiment of the invention the learning area is based on the immediate vicinity of the bases limited. This takes place against the background that operating points, which are relatively far from the base and be driven stationary for a long time, based on the base can be learned incorrectly. The narrowed down Learning areas need for quick learning too functionally important points that are driven as often as possible will be laid.

Operating points that are driven over a long period of time are stationary not learned after a set time.

It is particularly advantageous if the correction level is limited becomes. This means that there is a threshold for the amount the correction values K1, K2, K3 can be predetermined. Furthermore can be provided that the gradient, that is the Slope of the plane is limited. This means that the Difference between two correction values a threshold must not exceed. This limitation protects against incorrect ones Extrapolations e.g. after the start.

It is particularly advantageous if these bases with the largest possible distance from each other.

A particularly advantageous embodiment results if several correction levels can be specified. Particularly advantageous it is when for the three functional areas (Exhaust gas recirculation, full load and torque limitation) each a partial correction level can be specified. The number of levels can take any value. Due to the increased number of Sub-levels result in more bases and thus a larger one Flexibility, especially in the peripheral areas. Both No jumps may occur between the sub-levels. A line of intersection is preferably between the planes defined or there is a minimum and / or Maximum selection between two levels, or it is done an averaging of the level points at the operating point.

In a particularly advantageous embodiment, a Possible to reduce the influence of sensor errors. at the lambda probe is more accurate and at low lambda values The air mass meter is more accurate for large lambda values. from that is closed, for example, when the lambda full load is regulated, that an existing averaged difference between the two Air signals for the most part on a sensor error of the Air mass meter is based. The mean of this deviation when the full load is adjusted, global correction is possible of the air mass sensor. The mean of the deviation at regulated exhaust gas recirculation, for example when idling enables global correction of the lambda probe.

It is particularly advantageous if the speed values N1 and N2 are selected to be the same for the first and second operating points. The quantity values MK2 and MK3 are selected accordingly. The correction plane can be calculated very easily with these pairs of identical coordinates. The correction amount K of an operating point defined by the values MK and N results according to the following equation: K = K 1 + K 2 K 1 MK 2 MK 1 * ( MK - MK 1) + K 3 K 2 N 3 N 2 ( N - N 1)

In Figure 3, the work areas 1, 2 and 3 are for the different Sub-levels by means of a solid or a dash-dotted line separated. With crosses are the points at which the correction values Kn be determined. A smooth transition between work areas 2 and 3 and between work areas To achieve 2 and 1 will be one Straight line defined. The transition from sub-level 3 to sub-level 1 takes place in the area of the dash-dotted line Line through a minimal selection. There will be the correction amounts of the two levels and the smaller of the two Correction quantity is used.

It is preferably provided that the bases of the second Work area on the intersection line between the second and third correction plane and the intersection line between the second and first correction levels. Here lies a base on the intersection of the two intersection lines. Thus correction levels 1 and 2 as well as correction levels have 2 and 3 each have two common bases of the respective intersection line. Within the 1 and 3 Correction level is another base

Claims (10)

1. Method for controlling an internal combustion engine, in which an actuation signal is stored in a characteristic diagram as a function of at least two operating parameters, in that correction values for correcting the characteristic diagram can be obtained at at least three operating points, it being possible to determine the correction values on the basis of the deviation between a desired value and an actual value of an operating characteristic variable, characterized in that at least one correction plane is defined by means of at least three operating points and the assigned correction values and points of this correction plane are used as correction variables for actuation signals read out of the characteristic diagram.
2. Method according to Claim 1, characterized in that the fuel quantity signal is used as operating characteristic variable, it being possible to determine the actual fuel quantity on the basis of an air quantity signal and a lambda signal (λ).
3. Method according to Claim 1 or 2, characterized in that the characteristic diagram is the pump characteristic diagram of an auto-ignition internal combustion engine in which an actuation signal for a quantity-determining actuator is stored as a function of a rotational speed signal and a fuel quantity signal.
4. Method according to one of the preceding claims, characterized in that at least three operating ranges are provided, and in that in each operating range there is at least one operating point at which a correction value is determined.
5. Method according to Claim 4, characterized in that a correction plane can be predefined for each operating range.
6. Method according to one of Claims 4 or 5, characterized in that there is a first operating range at low rotational speeds and with small fuel quantities.
7. Method according to one of Claims 4 to 6, characterized in that there is a second operating range at low rotational speeds and with large fuel quantities.
8. Method according to one of Claims 4 to 7, characterized in that there is a third operating range at high rotational speeds and with large fuel quantities.
9. Method according to one of the preceding claims, characterized in that the correction values and/or the correction variables can be limited.
10. Device for controlling an internal combustion engine, in which an actuation signal is stored in a characteristic diagram as a function of at least two operating parameters, having means which obtain, at at least three operating points, correction values for correcting the characteristic diagram, it being possible to determine the correction values on the basis of the deviation between a desired value and an actual value of an operating characteristic variable, characterized in that means are provided which, by means of at least three operating points and the assigned correction values, define at least one correction plane and points of this correction plane are used as correction variables for actuation signals read out of the characteristic diagram.

Images