Method for determining optimized fuel injection history

Abstract

In a method for determining an optimized fuel injection profile in an internal combustion engine, a setpoint combustion profile is firstly defined. Furthermore, at least one influential parameter which influences the setpoint combustion profile is determined. With the influential parameter, a corrected fuel injection profile is determined in a closed-loop control process. This method is preferably repeated iteratively.

Claims

1. A closed-loop control feedback method for determining an optimized fuel injection profile of fuel in an internal combustion engine, comprising the steps of: determining a current injection profile; determining a setpoint combustion profile that extends over an entire combustion cycle, determining at least two influential parameters that influence the setpoint combustion profile from the group consisting of: a cylinder pressure gradient of a cylinder pressure increase at the start of combustion, a cylinder peak pressure, an indicated mean effective pressure, and a start of combustion, determining an actual value for each of the at least two determined influential parameters and at least two manipulated variables, one of the at least two manipulated variables corresponds to one of the at least two determined influential parameters and the other one of the at least two manipulated variables corresponds to the other one of the at least two determined influential parameters, each one of the at least two manipulated variables that correspond to the at least two determined influential parameters, respectively, is a scaling factor that is a continuously adjustable value and is used for minimizing a difference in a value between each of the at least two determined influential parameters and the actual value of the respective influential parameter; determining a corrected fuel injection profile in the closed-loop control based on the determined at least two influential parameters and the difference in value between each of the at least two determined influential parameters and the actual value of the respective influential parameter is a controlled variable, generating a corrected fuel injection profile using the determined at least two manipulated variables and the controlled variable for each one of the at least two influential parameters to reduce the value deviation in value between each of the at least two determined influential parameters and the actual value of the respective influential parameter, controlling the current injection profile using the generated corrected fuel injection profile to adjust the current injection profile, and repeating iteratively the preceding steps, wherein the scaling factor is determined by a closed-loop controller to reduce the value deviation between each of the at least two determined influential parameters and the actual value of the respective influential parameter for the current injection profile during the iterative repetition of generating the corrected fuel injection profile in the closed-loop control and adjusting the current injection profile.

2. The method according to claim 1, further comprising the steps of: determining a setpoint heat release from the setpoint combustion profile, and determining the corrected fuel injection profile by taking into consideration the setpoint heat release.

3. The method according to claim 1 further comprising: using at least two control loops for the optimization of the fuel injection profile, wherein the control loops are optimized in parallel.

4. The method according to claim 1 further comprising: optimizing the fuel injection profile with the aid of a first parameter for a first crankshaft angle range and performing optimization of the fuel injection profile with the aid of a second, different parameter, for a second crankshaft angle range, wherein the first and second crankshaft angle ranges do not overlap.

5. The method according to claim 1 further comprising: optimizing the fuel injection profile for a third crankshaft angle range with the aid of a third parameter, and optimizing the fuel injection profile for a further crankshaft angle range with the aid of a fourth parameter which differs from the third parameter, wherein the third and fourth crankshaft angle ranges overlap.

6. The method according to claim 1 wherein optimizing the fuel injection profile is performed with the closed-loop control of the cylinder peak pressure, and expanding a range of action of the closed-loop controller for the cylinder peak pressure toward earlier crankshaft angle positions until such time as the control deviation of the cylinder peak pressure is eliminated in the event that no scaling factor of the cylinder peak pressure can be determined such that a setpoint cylinder pressure is not overshot.

7. The method according to claim 1 further comprising: feeding the corrected fuel injection profile to a physically existing internal combustion engine, on which measured values of a resulting combustion profile are recorded.

8. The method according to claim 1 further comprising: feeding the corrected fuel injection profile to a simulated internal combustion engine, which obtains actual values of the combustion simulated therewith.

Description

DESCRIPTION OF THE DRAWINGS

(1) The invention will be described by way of example below on the basis of the figures, in which:

(2) FIG. 1 shows a setpoint combustion process on the basis of a setpoint cylinder pressure profile, taking into consideration the combustion characteristics of a start of combustion SOC, of a cylinder pressure increase a, of a cylinder peak pressure pMax and of an indicated mean effective pressure pmi,

(3) FIG. 2 is a schematic illustration of a closed-loop a controller of the cylinder pressure gradient, and the operating principle thereof with regard to an optimization of a fuel injection profile (EMK),

(4) FIG. 3 is a schematic illustration of a closed-loop pMax controller, and the operating principle thereof with regard to the EMK adjustment,

(5) FIG. 4 is a schematic illustration of a closed-loop pmi controller, and the operating principle thereof with regard to the EMK adjustment, and

(6) FIG. 5 is a schematic illustration of a closed-loop SOC controller, and the operating principle thereof with regard to the scaling of the ignition delay.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(7) A setpoint combustion profile is shown in FIG. 1. Here, in the lower part of FIG. 1, a setpoint cylinder pressure profile is taken as a starting point. This is determined by the following characteristics: the indicated mean effective pressure pmi, the start of combustion SOC, the cylinder pressure increase gradient a and the cylinder peak pressure pMax. These variables are shown in the lower half of the diagram. Accordingly, the cylinder pressure increases as the piston rises, until the start of combustion SOC begins at a crankshaft angle of approximately −2°. Thereafter, the pressure in the cylinder increases uniformly with the cylinder pressure increase gradient a, until it has reached the admissible maximum value pMax. From this point in time onward, it is sought to control the combustion in closed-loop fashion such that the pressure increases no further. The setpoint line of the cylinder pressure is thus the solid dark line of the cylinder pressure which transitions into the dotted horizontal line of pMax. By means of a pressure profile analysis, it is possible to generate further thermodynamic characteristic variables, such as for example the cylinder pressure profile, the heat release or the oxygen concentration profile of the gas enclosed in the cylinder. With the at least one cylinder pressure profile, but in particular also with these other thermodynamic state variables, which are illustrated in FIG. 1 versus the crankshaft angle, it is possible for an ignition delay profile to be calculated using a corresponding model. The indicated mean effective pressure of the high-pressure loop pmi is obtained by means of a closed-loop integral of the cylinder pressure profile over one combustion cycle. The heat release emerges from an energy balance and can be back-calculated using the pressure profile.

(8) Below, the fuel injection profile that is sought will be inferred from the described setpoint combustion profile. Here, as per FIG. 2, consideration is given to a correlation between the setpoint heat release profile CHRR (=cumulative heat release rate) and a so-called convertible fuel injection profile EMK taking into consideration the lower calorific value h.sub.α and a scaled factor k:

(9) $\begin{array}{cc}\mathrm{EMK}\approx \frac{1}{h_u}.Math.CHRR.Math.k& \mathrm{Equation}1\end{array}$

(10) The fuel injection profile, or fuel injection mass profile EMK, includes the fuel fragments that are available for the direct heat release in the combustion chamber. Here, consideration is given in particular to the cumulative injection mass. By contrast, for the control of the fuel injection, a hydraulic fuel injection profile EMH is required, which has a forward time offset in relation to the fuel injection profile EMK. The time offset corresponds here to the local ignition delay, which is known from the ignition delay model that is used.

(11) With these relationships, the following procedure for determining the fuel injection profile that is sought can be defined in the following steps:

Step 1

(12) Firstly, a definition of a setpoint combustion profile is performed. Here, it is for example possible to determine a setpoint cylinder pressure profile in the form of a so-called “Alpha process”. In the Alpha process, an increase gradient α of the pressure in the combustion chamber after the start of combustion SOC is defined. The increase gradient a may in particular be regarded as being constant until the maximum pressure pMax is reached. This combustion process is characterized significantly by the following combustion characteristics: start of combustion SOC, cylinder pressure gradient a and cylinder peak pressure pMax, as shown in FIG. 1. From this setpoint cylinder pressure profile, it is ultimately possible to derive thermodynamic state variable profiles (temperature, oxygen concentration, heat release and ignition delay) by means of a pressure profile analysis.

Step 2

(13) For the closed-loop control of the cylinder pressure gradient a, the setpoint heat release CHRR is scaled in accordance with equation 1. Here, particular attention must be paid to the determination of the scaling factor k. A starting value for k.sub.α,Initial is determined by means of a characteristic-map-based or model-based method as a function of the engine operating point, operating conditions and of the setpoint combustion profile, in this case in particular the setpoint value for the cylinder pressure gradient α.sub.soll, in the sense of pilot control. This pilot control value k.sub.α,Initial may self-evidently lead to a control deviation, for which reason, in further closed-loop control cycles, correction is performed, and k is obtained. This is realized for example by means of a linear integral closed-loop controller. The α control deviation is thus minimized by means of a correction of k.sub.α,kor in the sense of feedback-type closed-loop control. In the case of the closed-loop α controller, therefore, the cylinder pressure gradient a is the controlled variable, and the scaling factor k.sub.α is the manipulated variable. Alternatively and/or in addition to this, the pMax control deviation may be minimized by means of a corresponding correction of k.sub.pMax,kor. This is performed similarly to the closed-loop α control.

(14) Thus, here, the cylinder peak pressure pMax is the controlled variable and the corresponding scaling factor k.sub.pMax is the manipulated variable. The actual values α.sub.ist and pMax.sub.ist can be determined from the measured cylinder pressure profile of the preceding combustion profile. The use of both parameters results in two individual closed-loop controllers which are, in principle, independent, specifically: 1: closed-loop α controller and 2: closed-loop pMax controller, which are however based on the same principle, specifically equation 1. Thus, in the case of both controlled variables being used, the distinction between two different scaling factors is necessary, specifically one scaling factor for the closed-loop α controller, k.sub.α, and one for the closed-loop controller, k.sub.pMax. The respective scaling factors change the shape of the EMK profile in accordance with equation 1. The closed-loop control principles of the closed-loop α controller and of the closed-loop pMax controller, and the effects thereof on the EMK profile, are illustrated, for further clarification, in FIGS. 2 and 3. FIG. 2 shows, in the lower region, multiple optimization loops. Here, a first EMK.sub.Base profile is taken as a starting point, and the EMK value is optimized in the following loops, which are all denoted by EMK.sub.kor1,α. The resulting fuel injection profile EMK will be discussed in more detail in the next step 3:

Step 3

(15) With the known scaling factors, it is thus possible for the EMK profile to be composed in accordance with the described principle of equation 1. Here, the different areas of action of the closed-loop α controller and of the closed-loop pMax controller will differ. These areas of action are defined in particular by ranges of the crankshaft angle, and may in this case preferably not overlap. In particular, they adjoin one another. An example of this is shown in FIG. 1, wherein the optimization is performed by means of the closed-loop α controller in the crankshaft angle from −2° to 10°, and optimization is performed using pMax proceeding from the crankshaft angle of 10°. In the range between SOC and the onset of the pressure limitation, the scaling factor of the closed-loop a controller, k.sub.α, remains valid. From the onset of the peak pressure limitation, it is thus then the case that the scaling factor of the closed-loop pMax controller, k.sub.pMax, applies. The crankshaft angle position where the peak pressure pMax takes effect is known from the definition of the setpoint cylinder pressure profile already before the first closed-loop control cycle. If the pMax limitation is always exceeded by the actual value, that is to say the closed-loop pMax controller finds no scaling factor k.sub.pMax that leads to the peak cylinder pressure being adhered to, then the range of action of the closed-loop pMax controller (that is to say the range in which the CHRR is scaled by means of k.sub.pMax) is shifted gradually and automatically further in the direction of earlier crankshaft angle positions, until the pMax control deviation can be eliminated. The effect of the closed-loop α controller and of the closed-loop pMax controller on the EMK profile is schematically illustrated in FIGS. 2 and 3. Here, exemplary correction sequences of the scaling factors, with corresponding effects on the EMK profiles, are also illustrated. Correspondingly, in the lower part of FIG. 3, it is shown that, at the start, the profile is gradually shifted from EMK.sub.Base at a crankshaft angle shortly before 10° and in the individual corrective cycles EMK.sub.kor1,pMax, EMK.sub.kor2,pMax and EMK.sub.kor3,pMax.

(16) It is to be noted that, in a preferred embodiment, the maximum value of the EMK (EMK.sub.total) is determined neither by means of the closed-loop α controller nor by means of the closed-loop pMax controller. Instead, the EMK maximum value is determined by means of an additional closed-loop pmi controller. Here, “pmi” describes the indicated mean effective pressure of the high-pressure loop of the combustion process. The controlled variable of the closed-loop pmi controller is thus the indicated mean effective pressure, and the manipulated variable is the EMK maximum value. Therefore, an entire injection profile EMK.sub.total is corrected by means of the closed-loop pmi controller until such time as the pmi control deviation is sufficiently small. The EMK profile thus follows the scaled setpoint heat release profile CHRR within the closed-loop a control range or within the closed-loop pMax control range until EMK.sub.total is reached. The described closed-loop pmi controller is in this case made up of a pilot controller (for example characteristic-map-based or model-based) and an additional feedback-type closed-loop controller (for example integral closed-loop controller). A schematic illustration of the closed-loop pmi controller is shown in FIG. 4.

Step 4

(17) After an EMK profile resolved for a crankshaft angle has been determined in the preceding step, an associated EMH profile is determined in this step. For this purpose, each EMK fragment is shifted, by the local ignition delay, in the direction of earlier crankshaft positions, which ultimately leads to the generation of the desired EMH profile. As mentioned above, the SOC (=start of combustion) is a further important characteristic of the combustion process, which has already been taken into consideration in the creation of the setpoint cylinder pressure profile (step 1). For the correct closed-loop control of the SOC, use is therefore made of a further, independent closed-loop controller (closed-loop SOC controller). The closed-loop SOC controller thus minimizes the SOC control deviation through the use of a pilot controller (for example characteristic-map-based or model-based) and an integral feedback-type closed-loop controller component. The pilot control may, in the usage situation described, be described by means of an ignition delay model which has already been used in step 1 for the generation of the setpoint state variable profile. The actual SOC is, analogously to a and pMax, read out from the measured cylinder pressure profile of the preceding working cycle. As manipulated variable of the closed-loop SOC controller, the ignition delay is used, which, in accordance with the SOC deviation, is adjusted with a constant factor fak.sub.Zv. By means of the scaling of the ignition delay, the horizontal interval (crankshaft angle difference) between EMH and EMK is corrected, such that the SOC control deviation is sufficiently minimized. To illustrate the operating principle, the closed-loop SOC controller is schematically illustrated in FIG. 5. Here, an exemplary correction sequence of the ignition delay correction factor, with the corresponding effect on the EMK, is illustrated.

Step 5

(18) After the EMH profile has been determined in the preceding step, the EMH profile must then be converted, by means of a so-called digitalization method taking into consideration the fuel injector specifications, into a digital fuel injection profile. Here, from the hydraulic setpoint injection profile, an electrical injection profile is inferred, because it is only in this way that the injection profile can be realized by means of a commercial injector.

(19) Particular advantages resulting from the above-mentioned closed-loop control are thus a concept for closed-loop combustion control, which combines various closed-loop sub-controllers with one another to form a closed overall concept (closed-loop α/pMax/pmi/SOC control), which furthermore requires no further closed-loop controllers. Likewise, the concept permits, for the first time, the synchronous closed-loop control of the relevant combustion characteristics, specifically: peak pressure gradient a, peak pressure pMax, indicated mean effective pressure of the high-pressure loop pmi, start of combustion SOC.

(20) It is to be emphasized here that the closed-loop sub-controllers collectively form a convertible injection mass profile EMK. Here, mutually independent ranges of the EMK are addressed by the respective closed-loop sub-controllers. For a predefined setpoint combustion profile, the overall closed-loop control concept thus identifies a unique EMK profile, which also leads to a unique fuel injection profile. The EMK is thus used as a superordinate manipulated variable of the closed-loop control concept. It is to be highlighted here that, in this way, in particular, the parallel closed-loop control of the characteristics a and pMax is made possible by virtue of the closed-loop controller automatically determining the respective injection quantities, the respective injection times and the injection pattern (number of injections). The closed-loop control concept may be used both for transient closed-loop control methods (for example online implementation of the algorithms on the production control unit) or as a tool for targeted thermodynamic engine calibration of the fuel injection profile (for example in an experimental engine test environment). The fuel injection profiles calculated by means of the closed-loop controller can finally be stored by means of the characteristic maps of the production control unit and replicated in practical operation. Here, considerable calibration effort can be saved in relation to conventional calibration methods.

REFERENCE DESIGNATIONS

(21) CHRR Setpoint heat release

(22) EMK Fuel injection profile

(23) α Cylinder pressure gradient, cylinder pressure increase

(24) pmi Indicated mean effective pressure

(25) SOC Start of combustion

(26) pMax Cylinder peak pressure

(27) k Various scaling factors