Method for operating an internal combustion engine performing an injection quantity correction

Abstract

A method for operating a combustion engine performing an injection quantity correction is described. A total injection quantity per pulse of an injector is divided into a plurality of smaller equal quantity pulses. The smaller quantity pulses are implemented in ballistic injector mode. On the basis of this step, a corresponding offset correction is carried out. After the offset correction has been applied, a further correction is carried out in linear injector mode. An additional alternative for performing an injection quantity correction without additional sensor hardware is thereby provided.

Claims

1. A method for operating a combustion engine having an injector, the method comprising: dividing a total injection quantity X per pulse of the injector into a plurality of smaller equal quantity pulses X′ of equal total quantity, the smaller quantity pulses are implemented in ballistic injector mode; operating the smaller quantity pulses X′ with an activation duration t′ according to a nominal characteristic curve of the injector; back-calculating to an actual injection quantity X″ from an air/fuel ratio; searching for the actual injection quantity X″ on the nominal characteristic curve in order to determine the nominal necessary activation time t″; and determining a difference between the activation duration t′ and the nominal necessary activation time t″; and performing a corresponding offset correction of the injection quantity supplied by the injector; wherein after the offset correction has been applied, the method further includes: dividing a total injection quantity Y per pulse of the injector into a plurality of smaller equal quantity pulses Y′, wherein the smaller quantity pulses are implemented in linear injector mode; operating the smaller quantity pulses Y′ with an activation duration s′ according to a nominal characteristic curve; obtaining a quantity Y″; determining a deviation between the smaller quantity pulses Y′ and the obtained quantity Y″ as the deviation of the gradient of the nominal characteristic curves at that point; and repeating this procedure for various quantities/activation times and calculating and correcting the deviation of the gradient for the entire linear characteristic curve.

2. The method as claimed in claim 1, wherein the method is performed for multiple different operating points.

3. The method as claimed in claim 2, wherein a frequency distribution of the determined corrections is evaluated in order to increase a detection precision.

4. The method as claimed in claim 1, further comprising: detecting, using a sensor, an opening of the injector, wherein the method is performed for the plausibilization of the detection by the sensors.

Description

DESCRIPTION OF DRAWINGS

(1) FIG. 1 is a schematic illustration of five opening phases of an injector;

(2) FIG. 2 shows a diagram which shows the flow through an injector in dependence on the needle stroke;

(3) FIG. 3 shows a diagram which shows the injection rate as a function of time of the injector;

(4) FIG. 4 shows a diagram which shows the injection quantity as a function of the control duration;

(5) FIG. 5 shows a diagram which shows the injection rate as a function of time on the left and the injection rate as a function of time for individual partial pulses on the right;

(6) FIG. 6 shows a diagram which shows the injection quantity as a function of the control duration;

(7) FIG. 7 shows a diagram which shows the injection quantity as a function of the control duration;

(8) FIG. 8 shows a diagram which shows the injection rate as a function of time for the total pulse on the left and the injection rate as a function of time for partial pulses on the right;

(9) FIG. 9 shows a diagram which shows the injection quantity as a function of the control duration; and

(10) FIG. 10 shows a diagram which shows the injection quantity as a function of the control duration.

(11) Like reference symbols in the various drawings indicate like elements.

DETAILED DESCRIPTION

(12) The present disclosure provides a method for operating a combustion engine, having at least one injector, performing an injection quantity correction. Such a method will be explained hereinbelow.

(13) The method utilizes the different injector properties in ballistic mode (needle stop or maximum flow is not reached=position A-C in FIG. 1) and in linear mode (needle stop or maximum flow is reliably reached, position D in FIG. 1). FIG. 1 shows various opening phases of an injector, where 1 denotes a nozzle needle, 2 denotes an armature, 3 denotes a magnet coil, 4 denotes a nozzle plate and 5 denotes a mechanical stop. FIG. 1 is merely a schematic illustration, where no springs, guides etc. are shown.

(14) For the sake of simplicity, only a decrease in the injection quantity (broken lines) relative to the normal quantity (solid lines) is shown in the following figures. However, all the following statements also apply analogously to an increase in the quantity.

(15) A fundamental wear/tolerance point is the clearance of the armature 2 (position A to B in FIG. 1). A changed clearance in principle represents a temporal offset of the quantity characteristic curve (FIG. 4) and at the injection rate start (B to B.sub.L in FIG. 3). In the illustration flow over needle stroke, a changed clearance is not noticeable (B to B.sub.L in FIG. 2).

(16) The further substantial wear/tolerance point is the maximum flow (position D in FIG. 1). The flow is changeable via tolerances/wear of the nozzle plate 4 (D1 in FIG. 1) or by tolerances/wear of the stop 5 (D2 in FIG. 1). A change in the maximum flow in principle represents a gradient change in the quantity characteristic curve (FIG. 4) and a reduced maximum rate (D to D.sub.L in FIG. 3). A changed flow is also visible in the illustration flow over needle stroke (D to D.sub.L in FIG. 2).

(17) The method utilizes this behavior. If the working region of the injector is in the linear region, the algorithm can be started. This is virtually already the case at very low engine loads, so that the detection can be applied virtually without limitations.

(18) The present total injection quantity X is divided into a number n of smaller equal quantities X′ of equal total quantity. It is important that the small quantity pulses are implemented in ballistic injector mode (FIGS. 5 and 6). As the activation duration for the ballistic pulses there is implemented the control duration t′, which is found for the desired quantity on the nominal curve. On account of the solely ballistic pulses, any flow error is ineffective (FIG. 5). The injection quantity that is then implemented is too small. Since the entire air path remains unchanged during the transition, the determined lambda deviation is a result solely of the fuel error. The actual injection quantity X″ may therefore also be back-calculated from the air/fuel ratio. If X″ is then sought on the nominal characteristic curve, the nominal necessary activation time t″ is obtained (FIG. 6). The difference between t′ and t″ is the required offset correction. When this correction is taken into account, the characteristic curve of the injector in question shifts towards the nominal characteristic curve (FIG. 7).

(19) According to the same principle, the linear region is then considered with the offset correction applied. Here too, the total pulse is divided into multiple partial pulses. However, it must thereby be ensured that the partial pulses also reach the maximum flow or the needle stop (FIG. 8). The quantity Y is divided into the quantity Y′ and again operated with the activation duration according to the nominal characteristic curve. However, the quantity Y″ is then actually established. The deviation between Y′ and Y″ is equivalent to the deviation of the gradient of the characteristic curves at that point.

(20) If the procedure is repeated for ≥2 different quantities/activation times, the deviation of the gradient can be calculated and corrected for the entire linear characteristic curve (FIG. 10). If the entire method is performed for multiple different operating points and the frequency distribution of the determined corrections is taken into account, the detection precision can be increased significantly.

(21) A number of implementations have been described. Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the disclosure. Accordingly, other implementations are within the scope of the following claims.